{"id":5847,"date":"2026-08-10T00:56:28","date_gmt":"2026-08-10T00:56:28","guid":{"rendered":"https:\/\/cog-ist.com\/?post_type=blog_content&p=5847"},"modified":"2026-08-10T12:14:11","modified_gmt":"2026-08-10T12:14:11","slug":"bilincli-yapay-zeka-ve-biyolojik-dogalcilik-anil-k-seth","status":"publish","type":"blog_content","link":"https:\/\/cog-ist.com\/en\/blog_content\/bilincli-yapay-zeka-ve-biyolojik-dogalcilik-anil-k-seth\/","title":{"rendered":"Bilin\u00e7li Yapay Zek\u00e2 ve Biyolojik Do\u011falc\u0131l\u0131k – Anil K. Seth"},"content":{"rendered":"\n

Makalenin \u00d6zg\u00fcn Ad\u0131: Conscious artificial intelligence and biological naturalism<\/a><\/p>\n\n\n\n

<\/div>\n\n\n\n

Bu \u00e7eviri, makalenin yay\u0131mlanm\u0131\u015f ve telif hakk\u0131yla korunan nihai versiyonu \u00fczerinden de\u011fil, yazar(lar) taraf\u0131ndan CC BY 4.0 lisans\u0131 ile a\u00e7\u0131k eri\u015fime sunulan preprint versiyonu esas al\u0131narak haz\u0131rlanm\u0131\u015ft\u0131r.<\/p>\n\n\n\n

<\/div>\n\n\n\n

\u00d6z<\/strong><\/p>\n\n\n\n

<\/div>\n\n\n\n

Yapay zek\u00e2 (YZ) ilerleyi\u015fini s\u00fcrd\u00fcr\u00fcrken, YZ sistemlerinin yaln\u0131zca zeki de\u011fil, ayn\u0131 zamanda bilin\u00e7li olup olamayaca\u011f\u0131n\u0131 sormak do\u011fald\u0131r. Bu makalede, insanlar\u0131n YZ\u2019nin bilin\u00e7lenebilece\u011fini d\u00fc\u015f\u00fcnmesi i\u00e7in hangi gerek\u00e7elere sahip oldu\u011funu ele alacak ve bizi yanl\u0131\u015f y\u00f6nlendiren baz\u0131 \u00f6nyarg\u0131lar\u0131 te\u015fhis edece\u011fim. Bilin\u00e7li YZ\u2019nin ger\u00e7ek\u00e7i bir ihtimal olabilmesi i\u00e7in ne gerekti\u011fini sorgularken i\u015flemlemenin [computation<\/em>] bilince yeterli bir temel sa\u011flad\u0131\u011f\u0131 varsay\u0131m\u0131na kar\u015f\u0131 \u00e7\u0131kaca\u011f\u0131m. Bunun kar\u015f\u0131s\u0131nda, bilincin, ya\u015fayan organizmalar olarak do\u011fam\u0131za ba\u011fl\u0131 oldu\u011funu ileri s\u00fcren bir biyolojik do\u011falc\u0131l\u0131k (nat\u00fcralizm) bi\u00e7imini savunaca\u011f\u0131m. YZ i\u00e7in \u00e7e\u015fitli senaryolar\u0131 ele al\u0131p mevcut gidi\u015fatlar a\u00e7\u0131s\u0131ndan ger\u00e7ek yapay bilincin pek ihtimal dahilinde olmad\u0131\u011f\u0131n\u0131, ancak YZ beyne ve\/veya ya\u015fama daha \u00e7ok benzedik\u00e7e bilin\u00e7li yapay zek\u00e2n\u0131n daha makul bir olas\u0131l\u0131k haline geldi\u011fini ileri s\u00fcrerek tart\u0131\u015fmay\u0131 sonu\u00e7land\u0131raca\u011f\u0131m. Son olarak, ger\u00e7ekten bilin\u00e7li olan ya da ikna edici bi\u00e7imde bilin\u00e7li g\u00f6r\u00fcnen YZ\u2019nin \u00f6n\u00fcm\u00fcze koydu\u011fu etik meseleleri inceleyece\u011fim. Zihinlerimizi makine yarat\u0131mlar\u0131m\u0131za fazla ucuza b\u0131rak\u0131rsak onlar\u0131 abartmakla kalmay\u0131p kendimizi de k\u00fc\u00e7\u00fcmsemi\u015f oluruz.<\/p>\n\n\n\n

<\/div>\n\n\n\n

1.0 Giri\u015f<\/strong><\/p>\n\n\n\n

<\/div>\n\n\n\n
\n

\u201cTeknoloji bize, hayat hakk\u0131nda bildiklerimizi unutturabilir<\/em>.\u201d [(Turkle, 2021), p. xii]<\/p>\n<\/blockquote>\n\n\n\n

Yapay zek\u00e2 (YZ) sistemleri yetkinlik ve kabiliyet kazand\u0131k\u00e7a, bu sistemlerin zeki olman\u0131n yan\u0131 s\u0131ra bilin\u00e7li hale gelip gelemeyecekleri (hatta \u015fimdiden bilin\u00e7li olup olmad\u0131klar\u0131) sorusu g\u00fcndeme gelmektedir. Bu soruya kesin bir yan\u0131t vermek bug\u00fcn i\u00e7in m\u00fcmk\u00fcn de\u011fildir, zira bilin\u00e7 i\u00e7in asgari yeterli ko\u015fullar\u0131n ne oldu\u011fu konusunda bir uzla\u015fma yoktur. Yine de YZ\u2019nin \u015fimdiden bilin\u00e7li olabilece\u011fi ya da bilin\u00e7li hale gelebilece\u011fi fikri yayg\u0131nd\u0131r<\/strong> (Blum & Blum, 2024; Butlin ve di\u011ferleri, 2023; Chalmers, 2023; Dehaene ve di\u011ferleri, 2017). Bilin\u00e7li YZ ihtimalini nas\u0131l d\u00fc\u015f\u00fcnd\u00fc\u011f\u00fcm\u00fcz \u00f6nemlidir. Hem YZ sistemlerinin kendileri a\u00e7\u0131s\u0131ndan hem de insanlar\u0131n ve toplumun, bilin\u00e7li olan ya da bilin\u00e7li g\u00f6r\u00fcnen sistemlerden nas\u0131l etkilendi\u011fi bak\u0131m\u0131ndan \u00f6nemlidir.<\/strong> Ayr\u0131ca toplumlar\u0131m\u0131zda nas\u0131l bir YZ istedi\u011fimize karar vermemiz a\u00e7\u0131s\u0131ndan da belirleyici \u00f6neme sahiptir.<\/p>\n\n\n\n

Bu makalede, bilin\u00e7li bir YZ yaratma olanaklar\u0131n\u0131 ve bilin\u00e7li olan ya da \u00f6yle g\u00f6r\u00fcnen YZ\u2019nin sak\u0131ncalar\u0131n\u0131 ele alaca\u011f\u0131m. \u0130nsanlar\u0131n, neden YZ\u2019nin bilin\u00e7li olmaya do\u011fru ilerledi\u011fini d\u00fc\u015f\u00fcnebildi\u011fini sorgulayarak ba\u015flayacak ve bu olas\u0131l\u0131\u011f\u0131 abartmam\u0131za yol a\u00e7abilen baz\u0131 psikolojik \u00f6nyarg\u0131lar\u0131 saptayaca\u011f\u0131m. Ard\u0131ndan, bilin\u00e7li YZ olana\u011f\u0131n\u0131 temellendiren baz\u0131 varsay\u0131mlar\u0131 inceleyece\u011fim ve (kabaca) i\u015flemlemenin bilin\u00e7 i\u00e7in yeterli oldu\u011funu \u00f6ne s\u00fcren g\u00f6r\u00fc\u015f olan i\u015flemlemeselci i\u015flevselcili\u011fe <\/em>[computational functionalism<\/em>] odaklanaca\u011f\u0131m. Tart\u0131\u015fma boyunca, YZ\u2019nin bilin\u00e7li oldu\u011fu ya da olabilece\u011fi konusunda ku\u015fku duyulmas\u0131 i\u00e7in gerek\u00e7eler sunmay\u0131 hedefliyorum. Ama\u00e7, bilin\u00e7li bir yap\u0131t [artefact<\/em>] in\u015fa etmenin imk\u00e2ns\u0131z oldu\u011funu g\u00f6stermek de\u011fildir. Ama\u00e7, bu t\u00fcr yap\u0131tlar\u0131n, biyolojik sistemlere, klasik dijital i\u015flemlemenin olanak tan\u0131d\u0131\u011f\u0131ndan daha fazla \u00f6l\u00e7\u00fcde yak\u0131n olmak zorunda kalabilece\u011fini g\u00f6stermektir.<\/strong><\/p>\n\n\n\n

Daha sonra, burada bilincin yaln\u0131zca ya\u015fayan sistemlere \u00f6zg\u00fc (ancak zorunlu olarak hepsinde bulunmayan) bir \u00f6zellik oldu\u011funu savunan g\u00f6r\u00fc\u015f olarak anla\u015f\u0131lan biyolojik do\u011falc\u0131l\u0131\u011fa<\/em> ge\u00e7ece\u011fim. \u201cYa\u015fam\u201d\u0131n karbon temelli olmak zorunda olmad\u0131\u011f\u0131n\u0131 kabul ederek ve hi\u00e7bir neo-vitalist b\u00fcy\u00fcye yaslanmadan, bu g\u00f6r\u00fc\u015f\u00fcn birka\u00e7 farkl\u0131 versiyonunu \u00f6zetleyip gerek\u00e7elendirece\u011fim.<\/strong><\/p>\n\n\n\n

Son olarak daha geni\u015f meseleleri tart\u0131\u015farak, bilin\u00e7li olan ya da \u00f6yle g\u00f6r\u00fcnen YZ sistemlerinin do\u011furdu\u011fu kayg\u0131lar\u0131 vurgulayaca\u011f\u0131m. Ger\u00e7ek yapay bilin\u00e7 ihtimali hala uzak kalsa bile, bilin\u00e7li olduklar\u0131 y\u00f6n\u00fcnde sars\u0131lmas\u0131 g\u00fc\u00e7 bir izlenim yaratan YZ sistemleriyle ilgili ciddi etik ve toplumsal kayg\u0131lar bulunmaktad\u0131r.<\/strong><\/p>\n\n\n\n

<\/div>\n\n\n\n

1.1 Zek\u00e2 ve bilin\u00e7<\/em><\/p>\n\n\n\n

<\/div>\n\n\n\n

Ba\u015flarken \u201cbilin\u00e7\u201d ve \u201czek\u00e2\u201d i\u00e7in ge\u00e7ici tan\u0131mlara [working definition<\/em>] ihtiyac\u0131m\u0131z var. \u0130ki terim i\u00e7in de \u00fczerinde uzla\u015f\u0131lm\u0131\u015f tan\u0131mlar yok. Yine de baz\u0131 temel kurallar koymak gerekiyor.<\/p>\n\n\n\n

Zek\u00e2y\u0131, bir sistemin ne yapabildi\u011fini<\/em> ifade edecek \u015fekilde ele al\u0131yorum.<\/strong> Geni\u015f bir taramadan s\u00fcz\u00fclm\u00fc\u015f kullan\u0131\u015fl\u0131 bir tan\u0131m \u015fudur: \u201cfarkl\u0131 \u00e7evrelerde geni\u015f bir yelpazede ama\u00e7lara ula\u015fma yetene\u011fi\u201d [(Legg & Hutter, 2007), s. 23]. Bu anlamda ve di\u011fer \u00e7o\u011fu anlamda zek\u00e2, bir sistemin i\u015flevsel yetenekleriyle ilgilidir.<\/strong> Ak\u0131l y\u00fcr\u00fctme, ama\u00e7 y\u00f6nelimli planlama ve dilsel yetkinlik zek\u00e2n\u0131n bile\u015fenleridir. Ancak bu yeterliklerden yoksun olan insan-d\u0131\u015f\u0131 hayvanlar yine de zeki davran\u0131\u015f sergileyebilir. Zek\u00e2n\u0131n daha genel y\u00f6nleri aras\u0131nda zorluklara esnek bi\u00e7imde kar\u015f\u0131l\u0131k verebilme ve tecr\u00fcbeden \u00f6\u011frenebilme say\u0131labilir.<\/strong> \u00c7\u00f6lde evine nas\u0131l gidece\u011fini esnek bi\u00e7imde \u00f6\u011frenen bir kar\u0131nca, denilebilir ki, zeki bir davran\u0131\u015f sergilemektedir (Simon, 1988).<\/p>\n\n\n\n

YZ sistemleri, zek\u00e2 ile ili\u015fkilendirilen i\u015flevsel \u00f6zellikler sergileyen yapay sistemlerdir. Aksi belirtilmedik\u00e7e, YZ sistemlerinin (ya da \u201ceyleyicilerinin\u201d [agents<\/em>]<\/strong>[1]<\/a>) yerle\u015fik mimariler \u00fczerinde (CPU, GPU, TPU vb.) ger\u00e7ekle\u015ftirilen dijital i\u015flemlemelerden olu\u015ftu\u011funu ve ger\u00e7ek ya da sanal (\u00f6rne\u011fin robotik) bedenler ve \u00e7evrelerle etkile\u015fimleri i\u00e7erebilece\u011fini s\u00f6yleyen standart g\u00f6r\u00fc\u015f\u00fc benimseyece\u011fim<\/strong> (Russell & Norvig, 2003).<\/strong> Yapay Genel <\/em>Zek\u00e2 (YGZ [AGI<\/em>]), insanlar\u0131n zeki oldu\u011fu varsay\u0131lan b\u00fct\u00fcn bak\u0131mlardan insan d\u00fczeyine ula\u015fm\u0131\u015f ya da onu a\u015fm\u0131\u015f (hen\u00fcz varsay\u0131msal) YZ sistemlerine verilen add\u0131r. Bu sistemler, belirli g\u00f6rev alanlar\u0131 ve e\u011fitim rejimlerinin \u00f6tesinde genelleme yapabildikleri i\u00e7in \u201cgenel\u201d olarak nitelenir (Goertzel & Pennachin, 2007).<\/p>\n\n\n\n

Bilin\u00e7, zek\u00e2 ile ayn\u0131 \u015fey de\u011fildir.<\/strong> Bilin\u00e7, en az\u0131ndan birincil anlamda, bir sistemin ne yapt\u0131\u011f\u0131yla ilgili de\u011fildir.<\/strong> Bilin\u00e7 i\u00e7in yararl\u0131 bir referans tan\u0131m Thomas Nagel taraf\u0131ndan sunulmu\u015ftur: \u201cBilin\u00e7li bir organizma i\u00e7in, o organizma gibi olmak diye bir \u015fey vard\u0131r\u201d (Nagel, 1974, s. 436). Ba\u015fka bir deyi\u015fle, bilin\u00e7li bir sistem olmak bir \u015fey \u201cgibi hissettirir\u201d, bilin\u00e7li bir deneyim ya\u015fanmaktad\u0131r. \u00d6te yandan bilin\u00e7siz bir sistem olmak, herhangi bir \u015fey gibi hissettirmez, nitekim herhangi bir bilin\u00e7li deneyimin ya\u015fanmas\u0131 s\u00f6z konusu de\u011fildir.<\/strong> Burada \u201chissetme\u201dnin [feeling<\/em>] duygusal i\u00e7erim ta\u015f\u0131mas\u0131 gerekmez. Herhangi bir t\u00fcr bilin\u00e7li deneyim yeterlidir. Bir yarasa olmak muhtemelen bir \u015fey gibi hissettirmektedir ve bir ta\u015f olmak muhtemelen hi\u00e7bir \u015fey gibi hissettirmemektedir. Bir \u015feyin bilin\u00e7li olup olmad\u0131\u011f\u0131na ili\u015fkin olgusal bir durum [fact-of-the-matter<\/em>] bulundu\u011funu ve bilin\u00e7li olman\u0131n toplumsal ya da dilsel uzla\u015fmayla belirlenmedi\u011fini kabul ediyorum<\/strong> (bkz. Shanahan, 2024).<\/strong><\/p>\n\n\n\n

“Bilin\u00e7, zek\u00e2 ile ayn\u0131 \u015fey de\u011fildir. Bilin\u00e7, en az\u0131ndan birincil anlamda, bir sistemin ne yapt\u0131\u011f\u0131yla ilgili de\u011fildir.”<\/p><\/blockquote><\/figure>\n\n\n\n

\u201cBilin\u00e7\u201d ile \u201cfark\u0131ndal\u0131k\u201d\u0131 e\u015fanlaml\u0131 kullan\u0131yorum. \u201cDuyumsama\u201d [sentience<\/em>] teriminden ka\u00e7\u0131n\u0131yorum, \u00e7\u00fcnk\u00fc kimi zaman bilin\u00e7\/fark\u0131ndal\u0131k ile e\u015fanlaml\u0131 kullan\u0131lsa bile kar\u0131\u015f\u0131kl\u0131k yaratmaya daha elveri\u015flidir. Baz\u0131lar\u0131 \u201cduyumsama\u201dy\u0131, bilin\u00e7li deneyimin varl\u0131\u011f\u0131n\u0131 ima etmeksizin salt duyarl\u0131l\u0131k [sensitivity<\/em>] ya da tepki verebilirlik [responsiveness<\/em>] anlam\u0131nda kullan\u0131yor (Kagan ve di\u011ferleri, 2022; Kagan ve di\u011ferleri, 2023). Ba\u015fkalar\u0131 ise \u201cduyumsama\u201dy\u0131, deneyimlerin bir anlamda \u201ciyi\u201d ya da \u201ck\u00f6t\u00fc\u201d hissettirdi\u011fi zorunlu olarak de\u011ferlilik [valence<\/em>] i\u00e7eren bir bilin\u00e7 t\u00fcr\u00fc olarak kullanmaktad\u0131r (Thompson, 2022).<\/p>\n\n\n\n

Bu tan\u0131mlara g\u00f6re zek\u00e2 ile bilin\u00e7 kavramsal olarak a\u00e7\u0131k\u00e7a ayr\u0131d\u0131r.<\/strong> \u0130lkece bir canl\u0131 ya da sistem, o canl\u0131 ya da sistem gibi olmak diye bir \u015fey olmadan da zeki davran\u0131\u015f sergileyebilir, bunun tersi de m\u00fcmk\u00fcnd\u00fcr. Ancak uygulamada bu \u00f6zellikler birbirinden ba\u011f\u0131ms\u0131z g\u00f6r\u00fcnmez<\/strong>. Bilincin zek\u00e2n\u0131n baz\u0131 t\u00fcrleri i\u00e7in yararl\u0131 veya dahas\u0131 gerekli dahi olmas\u0131 m\u00fcmk\u00fcnd\u00fcr ve bilincin baz\u0131 t\u00fcrlerinin de kar\u015f\u0131l\u0131k gelen zek\u00e2 t\u00fcrlerini gerektirmesi ihtimal dahilindedir.[2]<\/a><\/p>\n\n\n\n

<\/div>\n\n\n\n

2.0 YZ\u2019nin bilin\u00e7li olabilece\u011fini ya da bilin\u00e7lenebilece\u011fini neden d\u00fc\u015f\u00fcn\u00fcr\u00fcz?<\/strong><\/p>\n\n\n\n

<\/div>\n\n\n\n

Bilin\u00e7li YZ fikrinin k\u00f6kleri derindir. On alt\u0131nc\u0131 y\u00fczy\u0131l Yahudi folklorunda, Josef ya da Yoselle denilen kilden yap\u0131lm\u0131\u015f bir golem, b\u00fcy\u00fcl\u00fc s\u00f6zler okundu\u011funda bilin\u00e7 kazan\u0131r ve rahats\u0131z edici bir \u015fiddet sergiler. Daha yak\u0131n bir d\u00f6nem i\u00e7in bilimkurgu, yapay bilin\u00e7 fikrini ke\u015ffe \u00e7\u0131kan hayali yarat\u0131mlar\u0131n g\u00fcvenilir bir kayna\u011f\u0131 olmu\u015ftur. Stanley Kubrick\u2019in 2001: A Space Odyssey<\/em> filmindeki HAL (1968), Alex Garland\u2019\u0131n Ex Machina<\/em>\u2019s\u0131ndaki Ava (2014) ve Kazuo Ishiguro\u2019nun Klara and the Sun<\/em>\u2019\u0131ndaki Klara (2021) bunlara \u00f6rnektir.<\/p>\n\n\n\n

Bu \u00f6yk\u00fclerin her birinde yapay yarat\u0131mlar, yarat\u0131c\u0131lar\u0131 taraf\u0131ndan, g\u00f6r\u00fcn\u00fc\u015fleriyle olmasa bile (\u00f6rne\u011fin HAL), zihinleri bak\u0131m\u0131ndan insana benzer olacak \u015fekilde tasarlanm\u0131\u015ft\u0131r. Bu durum \u00fc\u00e7 psikolojik \u00f6nyarg\u0131n\u0131n bir bile\u015fimini a\u00e7\u0131\u011fa \u00e7\u0131karmaktad\u0131r. \u0130nsanmerkezcilik [anthropocentrism<\/em>], insan-istisnac\u0131l\u0131\u011f\u0131 [human exceptionalism<\/em>] ve insanbi\u00e7imcilik [anthropomorphism<\/em>]…<\/strong> Bu \u00f6nyarg\u0131lar, zek\u00e2 ile bilincin birbirine kar\u0131\u015ft\u0131r\u0131lmas\u0131na yol a\u00e7abilir, bu da bilin\u00e7li YZ\u2019nin olanakl\u0131\u011f\u0131na dair yersiz bir iyimserli\u011fi k\u00f6r\u00fckleyecektir.<\/strong><\/p>\n\n\n\n

\u0130nsanmerkezcilik, insan\u0131 merkeze koyma ve d\u00fcnyay\u0131 insani de\u011ferler ve deneyimler \u00fczerinden yorumlama e\u011filimidir. \u0130nsanlar\u0131 di\u011fer ya\u015fam bi\u00e7imlerinden farkl\u0131 ve \u00fcst\u00fcn g\u00f6rme e\u011filimi olan insan-istisnac\u0131l\u0131\u011f\u0131yla yak\u0131ndan ili\u015fkilidir. Bu ikisi bir araya geldi\u011finde, \u00f6zellikle ve hatta belki m\u00fcnhas\u0131ran insana \u00f6zg\u00fcym\u00fc\u015f gibi g\u00f6r\u00fcnen zek\u00e2 boyutlar\u0131n\u0131 (\u00f6rne\u011fin dil ve di\u011fer \u201cileri\u201d bili\u015fsel yetiler) yine insana \u00f6zg\u00fc oldu\u011fu varsay\u0131lan niteliklerle (\u00f6rne\u011fin bilin\u00e7) s\u0131k\u0131 s\u0131k\u0131ya ba\u011fl\u0131 sayma ayart\u0131s\u0131 do\u011furur.<\/strong> Biz insanlar bilin\u00e7li oldu\u011fumuzu biliriz (daha do\u011frusu bunu kendimiz i\u00e7in bilir, ba\u015fkalar\u0131 hakk\u0131nda ise \u00e7\u0131kar\u0131m yapar\u0131z), zeki oldu\u011fumuzu d\u00fc\u015f\u00fcn\u00fcr ve \u00f6zel oldu\u011fumuzu hissederiz. Bu y\u00fczden de zek\u00e2 ile bilincin birlikte oldu\u011funu varsayar\u0131z. Bu d\u00fc\u015f\u00fcnme bi\u00e7iminin k\u00f6kleri en az\u0131ndan Descartes\u2019a kadar g\u00f6t\u00fcr\u00fclebilir. Descartes i\u00e7in maddi olmayan, ussal zihin hem insanlara \u00f6nem ta\u015f\u0131yan t\u00fcrden bir bilin\u00e7 bah\u015feder hem de insanlar\u0131 di\u011fer hayvanlardan ay\u0131r\u0131r (Shugg, 1968).<\/p>\n\n\n\n

\u0130nsanbi\u00e7imcilik, insanmerkezcili\u011fin tamamlay\u0131c\u0131s\u0131d\u0131r, \u0130nsans\u0131 \u00f6zellikleri, bu \u00f6zelliklere (zorunlu olarak) sahip olmayan \u015feylere yans\u0131tma e\u011filimidir.<\/strong> \u00d6rne\u011fin, y\u00fcz\u00fc olan \u015feylere duygular atfetme olas\u0131l\u0131\u011f\u0131m\u0131z, y\u00fcz\u00fc olmayan \u015feylere g\u00f6re daha y\u00fcksektir (Alais ve di\u011ferleri, 2021). Bu e\u011filim, Dennett\u2019in \u201cy\u00f6nelimsel duru\u015f\u201dunda [intentional stance<\/em>] oldu\u011fu gibi (Dennett, 1987), bir sistemin davran\u0131\u015f\u0131n\u0131 a\u00e7\u0131klamam\u0131za ve \u00f6ng\u00f6rmemize yard\u0131mc\u0131 oldu\u011funda i\u015fe yarar, ancak temelsiz bi\u00e7imde insans\u0131 \u00f6zellikler atfedilmesi zay\u0131f a\u00e7\u0131klamalara ve\/veya g\u00fcvenilmez \u00f6ng\u00f6r\u00fclere yol a\u00e7t\u0131\u011f\u0131nda sorunlu olabilir. \u0130nsanmerkezcilik ve insan-istisnac\u0131l\u0131\u011f\u0131yla birle\u015fti\u011finde insanbi\u00e7imcilik, dil ve \u201cy\u00fcksek\u201d bili\u015f gibi insan i\u00e7in en ay\u0131rt edici sayd\u0131\u011f\u0131m\u0131z insana \u00f6zg\u00fc nitelikleri sergiliyor gibi g\u00f6r\u00fcnen \u015feylere bilin\u00e7 atfetme e\u011filimini besleyebilir.<\/strong><\/p>\n\n\n\n

Bu \u00f6nyarg\u0131lar, YZ\u2019nin bilince giden soylu bir yolda oldu\u011fu g\u00f6r\u00fc\u015f\u00fcn\u00fc bir \u00f6l\u00e7\u00fcde a\u00e7\u0131klar. \u00c7\u00fcnk\u00fc makineler ak\u0131lland\u0131k\u00e7a sonunda bilin\u00e7li olacaklar\u0131 bir e\u015fi\u011fi a\u015facaklar\u0131 d\u00fc\u015f\u00fcn\u00fcl\u00fcr. Sistem bir kez bize, (insanmerkezci bir \u015fekilde) insanlara \u00f6zg\u00fc oldu\u011funu hissetti\u011fimiz bir \u00f6zelli\u011fi kazanm\u0131\u015f gibi g\u00f6r\u00fcnd\u00fc\u011f\u00fcnde, ona (insanbi\u00e7imci bir \u015fekilde) bilin\u00e7 gibi di\u011fer ay\u0131rt edici say\u0131lan \u00f6zellikleri atfetme e\u011filimimiz artar.<\/strong> S\u00f6z konusu e\u015fik, \u00f6rt\u00fck ya da a\u00e7\u0131k bi\u00e7imde Yapay Genel Zek\u00e2\u2019ya ula\u015f\u0131lan nokta olarak varsay\u0131ld\u0131\u011f\u0131nda, insan-istisnac\u0131l\u0131\u011f\u0131n\u0131n katk\u0131s\u0131 daha da belirginle\u015fir. Dili, ay\u0131rt edici bir \u015fekilde insans\u0131 zek\u00e2n\u0131n kanonik ifadesi kaidesine yerle\u015ftirmemiz, mevcut b\u00fcy\u00fck dil modeli (LLM) dalgas\u0131n\u0131n sezgilerimizi neden bu kadar g\u00fc\u00e7l\u00fc bi\u00e7imde yoldan \u00e7\u0131kard\u0131\u011f\u0131n\u0131 da a\u00e7\u0131klar<\/strong> (Colombatto & Fleming, 2023; Floridi & Nobre, 2024; Mitchell, 2024).<\/strong><\/p>\n\n\n\n

LLM\u2019lerin yoldan \u00e7\u0131kar\u0131c\u0131 g\u00fcc\u00fcne \u00f6rnek olarak, yanl\u0131\u015f \u00e7\u0131kt\u0131 \u00fcrettiklerinde \u201chal\u00fcsinasyon g\u00f6r\u00fcyorlar\u201d demenin ne kadar s\u0131radanla\u015ft\u0131\u011f\u0131n\u0131 d\u00fc\u015f\u00fcnebiliriz. \u0130nsanlarda hal\u00fcsinasyonlar esasen bozulmu\u015f alg\u0131sal deneyimlerdir. Bu terimin LLM\u2019ler i\u00e7in kullan\u0131m\u0131, onlara \u00f6rt\u00fck bi\u00e7imde deneyimsel kapasiteler atfeder. Oysa \u201ckonfab\u00fclasyon yap\u0131yorlar\u201d demek daha yerinde olurdu. \u0130nsanlarda konfab\u00fclasyon, kas\u0131ts\u0131z bir \u015fekilde (ve \u00e7o\u011funlukla dilsel) i\u00e7erik uydurmay\u0131 ifade etmektedir, ki\u015fi uydurdu\u011funun fark\u0131nda de\u011fildir. Dolay\u0131s\u0131yla, \u00f6ncelikle yapmaya<\/em> ili\u015fkindir, deneyimlemeye<\/em> de\u011fil.<\/p>\n\n\n\n

Muhtemelen d\u00f6rd\u00fcnc\u00fc bir bile\u015fen daha vard\u0131r. Kendimizi b\u00fcy\u00fck bir uygarl\u0131k d\u00f6n\u00fc\u015f\u00fcm\u00fcn\u00fcn e\u015fi\u011finde g\u00f6rmeye d\u00f6n\u00fck teknoloji heveslisi [technophilic<\/em>] ayart\u0131…<\/strong> Bu g\u00f6r\u00fc\u015f, YZ\u2019nin anlay\u0131\u015f\u0131m\u0131z\u0131n ve denetimimizin \u00f6tesine kendi kendini h\u0131zla geli\u015ftirdi\u011fi Kurzweilci bir Tekillik d\u00fc\u015f\u00fcncesiyle betimlenir<\/strong> (Kurzweil, 2005, 2024).<\/strong> Bu ayart\u0131 k\u0131smen, \u00fcstel de\u011fi\u015fimin do\u011fas\u0131n\u0131 kavraman\u0131n psikolojik g\u00fc\u00e7l\u00fc\u011f\u00fcnden beslenir (Wagenaar & Sagaria, 1975). \u00dcstel bir e\u011frinin herhangi bir noktas\u0131ndan bak\u0131ld\u0131\u011f\u0131nda gelecek imk\u00e2ns\u0131zm\u0131\u015f\u00e7as\u0131na dikey do\u011frultuda g\u00f6r\u00fcn\u00fcr, ge\u00e7mi\u015f ise \u00f6nemsizmi\u015f\u00e7esine d\u00fczd\u00fcr (buna ra\u011fmen, \u00fcstel de\u011fi\u015fimi yine de k\u00fc\u00e7\u00fcmsemeye meyilliyizdir. Gelecekteki nesneler g\u00f6r\u00fcnd\u00fcklerinden daha yak\u0131nd\u0131r[3]<\/a>). Bir ba\u015fka etken de, insan-\u00f6tesi [posthuman<\/em>] arzunun, bilinci gelece\u011fin s\u00fcper bilgisayarlar\u0131n\u0131n el de\u011fmemi\u015f devrelerine y\u00fckleyerek biyolojik \u00f6l\u00fcml\u00fcl\u00fc\u011f\u00fcn s\u0131n\u0131rlar\u0131n\u0131 a\u015fma ve b\u00f6ylece sonsuza kadar ya\u015fama ya da en az\u0131ndan sonsuza dek \u201cvar olma\u201d iste\u011fidir.<\/strong> Hatta halihaz\u0131rda bir bilgisayar sim\u00fclasyonunun i\u00e7inde oldu\u011fumuz fikri dahi felsefi ilgi g\u00f6rm\u00fc\u015ft\u00fcr (Bostrom, 2003).<\/p>\n\n\n\n

Hepsini bir araya toplad\u0131\u011f\u0131m\u0131zda, YZ\u2019nin ka\u00e7\u0131n\u0131lmaz bi\u00e7imde ve belki de \u00e7ok yak\u0131nda ger\u00e7ek yapay bilince yol a\u00e7aca\u011f\u0131 g\u00f6r\u00fc\u015f\u00fcn\u00fcn yayg\u0131nla\u015fmas\u0131na \u015fa\u015fmamak gerekir. Ancak bu g\u00f6r\u00fc\u015f\u00fc besleyen baz\u0131 temel \u00f6nyarg\u0131lar\u0131 saptad\u0131ktan sonra, onlar\u0131 bilin\u00e7liymi\u015f gibi yarg\u0131lamaya daha yatk\u0131n h\u00e2le gelsek bile, makineleri daha zeki h\u00e2le getirmenin onlar\u0131 ka\u00e7\u0131n\u0131lmaz bi\u00e7imde bilin\u00e7li k\u0131laca\u011f\u0131n\u0131 varsaymaktan kurtuluruz.<\/strong><\/p>\n\n\n\n

Ger\u00e7ek yapay bilincin kendili\u011finden \u201cbedavaya\u201d gelece\u011fini varsaymak yerine, onu kas\u0131tl\u0131 olarak geli\u015ftirmek denenebilir. Bu y\u00f6nde bir strateji, bilince ili\u015fkin \u00e7e\u015fitli kuramlar\u0131n belirledi\u011fi i\u015flevsel ilkeleri<\/strong> (Butlin ve di\u011ferleri, 2023; Seth & Bayne, 2022) ya da bilincin farkl\u0131 boyutlar\u0131n\u0131 \u00e7\u00f6z\u00fcmlemekten t\u00fcreyen ilkeleri<\/strong> (Evers ve di\u011ferleri, 2025) YZ sistemlerine uygulamaktan ge\u00e7er.<\/strong> Bu stratejinin ba\u015far\u0131l\u0131 olabilmesi i\u00e7in bilincin, (dijital i\u015flemlemeye dayanan) YZ\u2019nin sahip olabilece\u011fi t\u00fcrden bir \u015fey olmas\u0131 gerekir.<\/strong> Tam da bu nokta bizi i\u015flemlemeselci i\u015flevselcili\u011fe<\/em> g\u00f6t\u00fcrmektedir.<\/p>\n\n\n\n

<\/div>\n\n\n\n

3.0 \u0130\u015flemleme, zihin ve bilin\u00e7<\/strong><\/p>\n\n\n\n

<\/div>\n\n\n\n

3.1 \u0130\u015flemlemeselci i\u015flevselcilik<\/em><\/p>\n\n\n\n

<\/div>\n\n\n\n

\u0130\u015flemlemeselci i\u015flevselcilik [computational functionalism<\/em>], belli t\u00fcrden i\u015flemlemelerin bilinci ger\u00e7ekle\u015ftirmek i\u00e7in yeterli oldu\u011fu g\u00f6r\u00fc\u015f\u00fcd\u00fcr.<\/strong> Bu g\u00f6r\u00fc\u015f, felsefede i\u015flevselcilik diye adland\u0131r\u0131lan fikirden t\u00fcretilmi\u015ftir. \u0130\u015flevselcilik, (kabaca) bilincin, bir sistemin neyden yap\u0131ld\u0131\u011f\u0131na de\u011fil ne yapt\u0131\u011f\u0131na ba\u011fl\u0131 oldu\u011funu s\u00f6yler<\/strong> (Putnam, 1975).<\/strong><\/p>\n\n\n\n

Daha belirgin olarak i\u015flevselcilik, bir sistemin uygun bir i\u015flevsel \u00f6rg\u00fctlenmeye<\/em> [functional organization<\/em>] sahip oldu\u011funda zihne sahip oldu\u011funu \u00f6ne s\u00fcrer<\/strong> (Putnam, 1967). \u0130\u015flevsel \u00f6rg\u00fctlenme, hem sistemin bir b\u00fct\u00fcn olarak girdi-\u00e7\u0131kt\u0131 e\u015flemelerini hem de nedensel yap\u0131s\u0131 d\u00e2hil olmak \u00fczere i\u00e7 \u00f6rg\u00fctlenmesini kapsar. Zihinsel durumlar, bu i\u015flevsel \u00f6rg\u00fctlenme i\u00e7inde roller \u00fcstlenir ve duyu girdileri, motor \u00e7\u0131kt\u0131lar ve di\u011fer zihinsel durumlarla etkile\u015fimleriyle karakterize edilir. Bilince uyguland\u0131\u011f\u0131nda i\u015flevselcilik, bilin\u00e7li zihinsel durumlar\u0131n, alt\u0131nda yatan d\u00fczeneklerin [mechanisms<\/em>] i\u015flevsel \u00f6rg\u00fctlenmesinin \u00f6zellikleri oldu\u011funu \u00f6ne s\u00fcrer. \u0130nsan s\u00f6z konusu oldu\u011funda bu d\u00fczenek (bedenlenmi\u015f ve g\u00f6m\u00fcl\u00fc [embodied and embedded<\/em>]) beyindir.<\/strong><\/p>\n\n\n\n

\u0130\u015flemlemeselci i\u015flevselcilik<\/em> ise, genel olarak zihin ve \u00f6zel olarak bilin\u00e7 i\u00e7in belirleyici olan i\u015flevsel \u00f6rg\u00fctlenmenin do\u011fas\u0131 gere\u011fi i\u015flemlemesel oldu\u011funu savunmaktad\u0131r<\/strong> (Putnam, 1988; Shagrir, 2010). Bu sav, neyin \u201ci\u015flemlemesel\u201d [computational<\/em>] say\u0131laca\u011f\u0131na ili\u015fkin can s\u0131k\u0131c\u0131 soruyu g\u00fcndeme getirir<\/strong> (N. G. Anderson & G. Piccinini, 2024; Cantwell Smith, 2002; Searle, 1990).<\/p>\n\n\n\n

\u0130\u015flemlemenin do\u011fas\u0131 u\u00e7suz bucaks\u0131z bir konudur, yine de s\u0131n\u0131rlar\u0131 kabaca da olsa \u00e7izmek faydal\u0131 olacakt\u0131r. Burada en ilgili kavram, Alan Turing\u2019in \u201cTuring makinesi\u201d fikri arac\u0131l\u0131\u011f\u0131yla tan\u0131tt\u0131\u011f\u0131 \u201cTuring i\u015flemlemesi\u201ddir<\/strong> (Turing, 1936). Bir Turing makinesi, sonlu say\u0131da kural\u0131 sonlu bir sembol k\u00fcmesine uygular ve b\u00f6ylece bir matematiksel fonksiyonun de\u011ferlerini i\u015flemler [compute<\/em>]. Ba\u015fka bir deyi\u015fle bir algoritma y\u00fcr\u00fct\u00fcr. Turing ayr\u0131ca her algoritma i\u00e7in bir Turing makinesi bulundu\u011funu \u00f6ne s\u00fcrer. Church\u2013Turing tezi olarak bilinen bu g\u00f6r\u00fc\u015f geni\u015f kabul g\u00f6rm\u00fc\u015ft\u00fcr. <\/strong>Turing, bunlar\u0131n yan\u0131s\u0131ra, yeterli depolama sa\u011fland\u0131\u011f\u0131nda herhangi bir belirli Turing makinesini ger\u00e7ekle\u015ftirebilen \u201cevrensel\u201d Turing makinesinin de tan\u0131m\u0131n\u0131 yapm\u0131\u015ft\u0131r (Copeland, 2024). Geleneksel YZ\u2019yi \u00e7al\u0131\u015ft\u0131ran dijital bilgisayarlar, evrensel Turing makinelerinin modern \u00f6rnekleridir.<\/p>\n\n\n\n

A\u015fa\u011f\u0131 yukar\u0131 ayn\u0131 d\u00f6nemde Warren McCulloch ve Walter Pitts, baz\u0131 Boole\u2019cu mant\u0131k i\u015flemlerinin herhangi bir bile\u015fiminin, epey sadele\u015ftirilmi\u015f yapay n\u00f6ron devreleriyle ger\u00e7ekle\u015ftirilebilece\u011fini g\u00f6sterdiler (McCulloch & Pitts, 1943). Daha sonra bu yapay sinir a\u011flar\u0131n\u0131n Turing makineleri olarak yorumlanabilece\u011fi ortaya konuldu (Kleene, 1956).[4]<\/a> Bu sonu\u00e7, biyolojik sinir sistemlerinin Turing anlam\u0131nda i\u015flemleme yapan makineler olarak d\u00fc\u015f\u00fcn\u00fclebilece\u011fi g\u00f6r\u00fc\u015f\u00fcn\u00fc desteklemi\u015ftir <\/strong>(Piccinini, 2023). Turing\u2019in \u00fcnl\u00fc \u201cMakineler d\u00fc\u015f\u00fcnebilir mi?\u201d sorusu <\/strong>(Turing, 1950) da bu g\u00f6r\u00fc\u015fe katk\u0131 sunmu\u015ftur.<\/strong>[5]<\/a><\/p>\n\n\n\n

Turing i\u015flemlemesi g\u00fc\u00e7l\u00fcd\u00fcr, ama her fonksiyon Turing-i\u015flemlenebilir de\u011fildir.<\/strong> Turing, Hilbert\u2019in ortaya koydu\u011fu \u201cdurma problemi\u201dne [halting problem<\/em>] verdi\u011fi yan\u0131tta bir i\u015flemlenemez [non-computable<\/em>] fonksiyon s\u0131n\u0131f\u0131n\u0131 bizzat tan\u0131mlam\u0131\u015ft\u0131r (Turing, 1936). Ba\u015fka \u00f6rnekler aras\u0131nda s\u00fcrekli de\u011fi\u015fkenler i\u00e7eren fonksiyonlar, stokastik\/rastlant\u0131sal \u00f6\u011feler bar\u0131nd\u0131ran fonksiyonlar ve ba\u015flang\u0131\u00e7 ko\u015fullar\u0131na s\u0131n\u0131rs\u0131z duyarl\u0131l\u0131k g\u00f6steren fonksiyonlar bulunur (\u00f6rne\u011fin deterministik kaos). Turing makinelerine dayanan dijital bilgisayarlar, bu fonksiyonlar\u0131n bir k\u0131sm\u0131n\u0131 sim\u00fcle edebilir ve yakla\u015f\u0131k olarak hesaplayabilir. Buna s\u00fcreklilik i\u00e7eren s\u00fcre\u00e7ler ile stokastik \u00f6\u011feler bar\u0131nd\u0131ran s\u00fcre\u00e7ler \u00f6rnek verilebilir (belirli tekil \u00f6rnekler yerine da\u011f\u0131l\u0131mlar d\u00fczeyinde). Nitekim bu durum i\u015flemlemesel modellemede s\u00fcrekli ya\u015fan\u0131r. Fakat bu kestirimler [approximations<\/em>] genellikle kesin de\u011fildir.[6]<\/a><\/p>\n\n\n\n

Turing i\u015flemlemesinin s\u0131n\u0131rl\u0131 kapsam\u0131, beyinler de d\u00e2hil olmak \u00fczere baz\u0131 sistemlerin Turing-d\u0131\u015f\u0131 i\u015flevler ger\u00e7ekle\u015ftirebilece\u011fi anlam\u0131na gelir. Zihinsel durumlar\u0131n (bilin\u00e7 d\u00e2hil) i\u015flemlemesel olmayan i\u015flevlere ba\u011fl\u0131 oldu\u011fu g\u00f6r\u00fc\u015f\u00fcne i\u015flemlemesel olmayan i\u015flevselcilik<\/em> denir<\/strong> (Piccinini, 2018, 2020). \u0130\u015flemlemesel olmayan n\u00f6ral i\u015flevler aras\u0131nda (s\u00fcre\u011fen) elektromanyetik alanlara ili\u015fkin s\u00fcre\u00e7ler, ince-taneli [fine-grained<\/em>] zamanlama ili\u015fkileri (Turing i\u015flemlemesinde belirleyici olan \u015fey s\u0131ralamad\u0131r, dinami\u011fin kendisi de\u011fil), serbest\u00e7e dif\u00fcze olan n\u00f6roiletkenler ve benzeri mekanizmalar say\u0131labilir. \u0130\u015flemlemesel olmayan biyolojik i\u015flevlere, zorunlu olarak belirli bir maddi \u00f6zelli\u011fi i\u00e7eren s\u00fcre\u00e7ler de girer. Sindirim, kan dola\u015f\u0131m\u0131 ve metabolizma buna \u00f6rnektir. \u00dcstelik i\u015flemlemesel i\u015flevlerle i\u015flemlemesel olmayan i\u015flevler bir arada bulunabilir. \u00d6rne\u011fin zihnin baz\u0131 y\u00f6nleri i\u015flemlemesel olabilirken, bilin\u00e7 \u00f6yle olmayabilir<\/strong> (Piccinini, 2023).<\/strong><\/p>\n\n\n\n

\u201c\u0130\u015flemleme\u201d kavram\u0131na ili\u015fkin daha pek \u00e7ok ba\u015fka tan\u0131m \u00f6nerilmi\u015ftir. Baz\u0131lar\u0131 Turing i\u015flemlemesinden daha dard\u0131r (\u00f6rne\u011fin bir ayg\u0131t\u0131n bir ki\u015fi taraf\u0131ndan belirli bir bi\u00e7imde kullan\u0131lmas\u0131n\u0131 gerektiren i\u015flemleme anlay\u0131\u015flar\u0131); ancak \u00e7o\u011fu daha geni\u015ftir (N. G. Anderson & G. Piccinini, 2024; Chalmers, 1996b; Kirkpatrick, 2022). Daha geni\u015f kapsaml\u0131 i\u015flemleme bi\u00e7imleri aras\u0131nda analog, sinirbi\u00e7imsel ve \u00f6l\u00fcml\u00fc i\u015flemleme [mortal computation<\/em>] say\u0131labilir. Bunlara ileride d\u00f6nece\u011fim. \u015eimdilik Turing i\u015flemlemesine odaklanmak yerindedir. \u00c7\u00fcnk\u00fc bu i\u015flemleme t\u00fcr\u00fc, ister yapay sinir a\u011flar\u0131na dayans\u0131n ister ba\u015fka y\u00f6ntemlere, geleneksel YZ\u2019nin temelindedir.<\/strong> Aksi belirtilmedik\u00e7e \u201ci\u015flemleme\u201d s\u00f6zc\u00fc\u011f\u00fcn\u00fc Turing i\u015flemlemesi anlam\u0131nda kullanaca\u011f\u0131m. (\u201cEnformasyon i\u015fleme\u201d [information processing<\/em>] ifadesi \u00e7o\u011fu zaman \u201ci\u015flemleme\u201d ile ayn\u0131 kapsamda kullan\u0131lsa da, bu ifade yarars\u0131z bi\u00e7imde mu\u011flakt\u0131r. Bunun bir nedeni enformasyon kuram\u0131n\u0131n betimleyici olarak neredeyse her \u015feye uygulanabilmesidir[7]<\/a>. Bir di\u011fer nedeni de \u201ci\u015fleme\u201dnin ne anlama geldi\u011finin s\u0131kl\u0131kla belirsiz olu\u015fudur.)<\/p>\n\n\n\n

Genel olarak i\u015flevselcilik ve \u00f6zellikle i\u015flemlemeselci i\u015flevselcilik, a\u00e7\u0131k ya da \u00f6rt\u00fck bi\u00e7imlerde geni\u015f kabul g\u00f6rmektedir. Bilin\u00e7li YZ\u2019ye ili\u015fkin ihtimalleri etkili bi\u00e7imde g\u00f6zden ge\u00e7iren bir \u00e7al\u0131\u015fmada Patrick Butlin, Robert Long ve \u00e7al\u0131\u015fma arkada\u015flar\u0131, kendilerini varsay\u0131m olarak i\u015flemlemeselci i\u015flevselcili\u011fi benimseyen kuram ve modellerle s\u0131n\u0131rland\u0131r\u0131r ve var\u0131lan sonu\u00e7lar\u0131n bu varsay\u0131m\u0131n ge\u00e7erlili\u011fine ba\u011fl\u0131 oldu\u011funu a\u00e7\u0131k\u00e7a belirtir (Butlin ve di\u011ferleri, 2023). Ancak her fonksiyon i\u015flemlenebilir olmad\u0131\u011f\u0131 ve her \u015fey i\u015flemleme ger\u00e7ekle\u015ftirmedi\u011fine g\u00f6re s\u00f6z konusu varsay\u0131m, sorgulanmaya de\u011ferdir.<\/strong><\/p>\n\n\n\n

<\/div>\n\n\n\n

3.2 \u0130\u015flemlemeselci i\u015flevselcili\u011fin cazibesi<\/em><\/p>\n\n\n\n

<\/div>\n\n\n\n

Nas\u0131l ki zek\u00e2 ile bilin\u00e7 aras\u0131ndaki ili\u015fkilere psikolojik ve tarihsel etkenler y\u00f6n veriyorsa, i\u015flemlemeselci i\u015flevselcili\u011fin \u00e7ekicili\u011fini a\u00e7\u0131klamaya yard\u0131mc\u0131 olan daha geni\u015f bir ba\u011flam da mevcuttur.<\/strong> Bu ba\u011flam\u0131n ay\u0131rd\u0131na varmak, s\u00f6z konusu g\u00f6r\u00fc\u015f\u00fcn ikna edicili\u011fini de\u011ferlendirmeye katk\u0131da bulunur.<\/p>\n\n\n\n

Genel olarak i\u015flevselcilik, bir \u00f6l\u00e7\u00fcde davran\u0131\u015f\u00e7\u0131l\u0131k, \u00f6zde\u015flik kuramlar\u0131 (zihinsel durumlar\u0131n fiziksel beyin durumlar\u0131na kat\u0131 indirgenmesini \u00f6ne s\u00fcren yakla\u015f\u0131mlar) ve ikicilik (d\u00fcalizm) gibi di\u011fer felsefi \u00e7er\u00e7evelerin yetersiz oldu\u011fu alg\u0131s\u0131ndan beslenmi\u015ftir<\/strong> (Shagrir, 2010).<\/strong> \u0130\u015flevsel \u00f6rg\u00fctlenmeye odaklanmak, davran\u0131\u015f\u00e7\u0131l\u0131k ile \u00f6zde\u015flik kuram\u0131 aras\u0131nda, ayn\u0131 zihinsel durumun farkl\u0131 yollarla ger\u00e7ekle\u015febilece\u011fini ifade eden \u00e7oklu ger\u00e7ekle\u015febilirlik<\/em> [multiple realisability<\/em>] fikrini bar\u0131nd\u0131ran cazip bir orta yol sunmu\u015ftur (Polger & Shapiro, 2016; Putnam, 1967). Ancak bir nokta \u00f6nemlidir ki \u00e7oklu ger\u00e7ekle\u015febilirlik, her zihinsel durumun herhangi bir fiziksel sistemde uygulanabilece\u011fi anlam\u0131na gelmez. Baz\u0131 fiziksel sistemler bu i\u015fi yapmaya elveri\u015fli olmayabilir.<\/strong><\/p>\n\n\n\n

\u0130\u015flemlemeselci i\u015flevselcili\u011fe y\u00f6nelik co\u015fku, belirli bir i\u015flemleme anlay\u0131\u015f\u0131 olan Turing makinesi ile beynin \u00e7ok katmanl\u0131, karma\u015f\u0131k biyolojik yap\u0131s\u0131n\u0131 \u00e7arp\u0131c\u0131 bi\u00e7imde sadele\u015ftiren yapay sinir a\u011f\u0131 aras\u0131nda kurulan matematiksel bir mant\u0131k evlili\u011fine kadar geri g\u00f6t\u00fcr\u00fclebilir.<\/strong> Bilgisayar bilimi ve yapay zek\u00e2n\u0131n erken bi\u00e7imleri 20. y\u00fczy\u0131l\u0131n ikinci yar\u0131s\u0131nda b\u00fcy\u00fck bir h\u0131zda ilerledik\u00e7e bu co\u015fku ivme kazanm\u0131\u015ft\u0131r (Boden, 2008; Dupuy, 2009). S\u00f6z konusu ilerlemeler, beynin bir bilgisayar oldu\u011fu metaforunu yerle\u015ftirmeye, beyinlerin ne yapt\u0131\u011f\u0131na dair enformasyon i\u015fleme dilini normalle\u015ftirmeye ve bili\u015fsel bilimi bir disiplin olarak g\u00fc\u00e7lendirmeye yard\u0131mc\u0131 olmu\u015ftur. Yine de metaforlar, nihayetinde, yaln\u0131zca metaforlard\u0131r<\/strong> (Cobb, 2020). Zihinsel s\u00fcre\u00e7lerin \u00e7o\u011fu zaman i\u015flemleme terimleriyle faydal\u0131 bi\u00e7imde tan\u0131mlanabilmesi, beynin ger\u00e7ekten i\u015flemleme yapt\u0131\u011f\u0131 ya da bilincin bir i\u015flemleme t\u00fcr\u00fc oldu\u011fu sonucunu \u00e7\u0131karmaya yetmez.<\/strong><\/p>\n\n\n\n

Geleneksel dijital sistemler i\u00e7in bilin\u00e7li YZ\u2019nin m\u00fcmk\u00fcn olabilmesi, ilgili i\u015flemlemelerin bilince yeterli bir temel sa\u011flamas\u0131n\u0131 gerektirir. Bunun ger\u00e7ekle\u015fmesi i\u00e7in hem i\u015flemlemeselci i\u015flevselcili\u011fin ge\u00e7erli olmas\u0131 hem de<\/em> YZ sistemlerinin s\u00f6z konusu i\u015flemlemeleri uygulayabilmesi gerekir. \u0130\u015flemlemeselci i\u015flevselcili\u011fin s\u0131k\u00e7a varsay\u0131l\u0131yor olmas\u0131, onun do\u011fru oldu\u011fu anlam\u0131na gelmez. Bilim yazar\u0131 Oshan Jarrow\u2019un dedi\u011fi gibi, \u201cayn\u0131 varsay\u0131m\u0131, ona geri d\u00f6n\u00fcp s\u0131namadan s\u00fcrd\u00fcr\u00fcrsek, sorunun kendisi yava\u015f yava\u015f ortadan kaybolmaya ba\u015flar.\u201d<\/strong>[8]<\/a><\/p>\n\n\n\n

<\/div>\n\n\n\n

3.3 Substrat ba\u011f\u0131ms\u0131zl\u0131\u011f\u0131, esneklik ve n\u00f6ral ikame<\/em><\/p>\n\n\n\n

<\/div>\n\n\n\n

Bilin\u00e7li YZ ile hesa i\u015flevselcilik aras\u0131ndaki ili\u015fkiyi ortaya koyman\u0131n bir yolu, meseleyi substrat ba\u011f\u0131ms\u0131zl\u0131\u011f\u0131 a\u00e7\u0131s\u0131ndan ele almaktad\u0131r. Substrat (ya da e\u015fde\u011fer bi\u00e7imde ortam) ba\u011f\u0131ms\u0131zl\u0131\u011f\u0131, \u00e7oklu ger\u00e7ekle\u015febilirlikle yak\u0131ndan ili\u015fkilidir. Bu g\u00f6r\u00fc\u015fe g\u00f6re, ayn\u0131 t\u00fcrden zihinsel durumlar (bilin\u00e7li durumlar d\u00e2hil), farkl\u0131 fiziksel substratlara sahip sistemlerde meydana gelebilir. Ba\u015fka bir deyi\u015fle, farkl\u0131 t\u00fcrden maddelerden olu\u015fan sistemlerde ortaya \u00e7\u0131kabilir. Ancak s\u0131n\u0131rs\u0131z substrat ba\u011f\u0131ms\u0131zl\u0131\u011f\u0131n\u0131n ge\u00e7erli olmas\u0131 pek olas\u0131 de\u011fildir.<\/strong> \u00c7oklu ger\u00e7ekle\u015febilirlikte oldu\u011fu gibi, baz\u0131 fiziksel substratlar bu i\u015fi yapmaya elveri\u015fli olmayabilir. Bu nedenle, substrat esnekli\u011fi <\/em>terimini kullanaca\u011f\u0131m. Buna g\u00f6re (bilin\u00e7li) zihinsel durumlar karbon temelli biyolojik substratlara s\u0131k\u0131 s\u0131k\u0131ya ba\u011fl\u0131 de\u011fildir, ba\u015fka t\u00fcrden maddi yap\u0131larda da ger\u00e7ekle\u015febilir, fakat bu her t\u00fcrl\u00fc maddi yap\u0131 i\u00e7in ge\u00e7erli olmayabilir<\/strong> (benzer bir fikir i\u00e7in bkz. Kirkpatrick, 2022).<\/strong>[9]<\/a><\/p>\n\n\n\n

\u0130\u015flemlemeselci i\u015flevselcilik, substrat esnekli\u011fi ile yak\u0131ndan ba\u011flant\u0131l\u0131d\u0131r. E\u011fer i\u015flemlemeselci i\u015flevselcilik do\u011fruysa, ilgili i\u015flemlemeleri ger\u00e7ekle\u015ftirebilen substratlar i\u00e7in belli bir substrat esnekli\u011fi sonucu do\u011far. Ba\u015fka bir ifadeyle, i\u015flemlemeselci i\u015flevselcilik, substrat esnekli\u011fi i\u00e7in yeterlidir. Fakat zorunlu de\u011fildir. Zihin ya da bilin\u00e7 hakk\u0131nda i\u015flemlemeselci kuramlar\u0131 benimsemeksizin de belli bir substrat esnekli\u011fi tasavvur edilebilir. Yine de substrat esnekli\u011fi belli \u00f6l\u00e7\u00fcde ge\u00e7erliyse, i\u015flemlemeselci i\u015flevselcilik destek kazan\u0131r. \u00c7\u00fcnk\u00fc bu g\u00f6r\u00fc\u015f, zihinsel durumlar\u0131, ilgili i\u015flemlemeleri ger\u00e7ekle\u015ftirebilme yetene\u011finin \u00f6tesinde, substrat\u0131n maddi \u00f6zelliklerine ba\u011f\u0131ml\u0131 k\u0131lmadan a\u00e7\u0131klar. Bilin\u00e7li YZ i\u00e7in, en az\u0131ndan tart\u0131\u015fma konusu substrat, yani silikon a\u00e7\u0131s\u0131ndan substrat esnekli\u011finin ge\u00e7erli olmas\u0131 gerekir.<\/strong><\/p>\n\n\n\n

Bu ba\u011flamda substrat esnekli\u011fine dair en bilinen arg\u00fcmanlardan biri \u201cn\u00f6ral ikame\u201d d\u00fc\u015f\u00fcnce deneyidir (Chalmers, 1995; Haugeland, 1980; Pylyshyn, 1980). Temel senaryoda, bir ki\u015finin beyin h\u00fccreleri[10]<\/a> tek tek silikon e\u015fde\u011ferleriyle de\u011fi\u015ftirilir. Her silikon beyin h\u00fccresi, biyolojik e\u015fleni\u011finin [counterpart<\/em>] girdi\/\u00e7\u0131kt\u0131 e\u015flemesini her bir durumda tam olarak taklit eder. Sonunda ki\u015finin beyninde yaln\u0131zca silikon par\u00e7alar kal\u0131r, buna kar\u015f\u0131n i\u015flevsel \u00f6rg\u00fctlenme b\u00fct\u00fcn\u00fcyle korunur ve d\u0131\u015far\u0131dan bak\u0131ld\u0131\u011f\u0131nda ki\u015fi eskisi gibi davranmay\u0131 s\u00fcrd\u00fcr\u00fcr. Tek bir beyin h\u00fccresinin de\u011fi\u015ftirilmesi ki\u015finin bilincinde fark yaratm\u0131yorsa (ki bu pek olas\u0131 g\u00f6r\u00fcnmez), neden y\u00fcz tanesinin ya da tamam\u0131n\u0131n de\u011fi\u015ftirilmesi fark yarats\u0131n? Bilin\u00e7 yava\u015f\u00e7a s\u00f6n\u00fcp gidecek midir?<\/strong> E\u011fer \u00f6yleyse, bilin\u00e7li deneyim t\u00fcm\u00fcyle kaybolurken davran\u0131\u015f\u0131n de\u011fi\u015fmeden kalmas\u0131 akla yatk\u0131n m\u0131d\u0131r? Bu tuhaf sonu\u00e7lardan ka\u00e7\u0131nman\u0131n yolu, silikonun substrat esnekli\u011fini kabul etmek gibi g\u00f6r\u00fcn\u00fcr.<\/strong><\/p>\n\n\n\n

Bu d\u00fc\u015f\u00fcnce deneyi \u00e7e\u015fitli y\u00f6nlerden ele\u015ftirilebilir. Block (2019), deneyi, ispatlamaya \u00e7al\u0131\u015ft\u0131\u011f\u0131 \u015feyi varsaymakla su\u00e7lar. E\u011fer bilin\u00e7 substrata ba\u011f\u0131ml\u0131ysa, n\u00f6ral ikame ilerledik\u00e7e ki\u015finin bilin\u00e7li deneyimine ger\u00e7ekten de bir \u015feyler olmal\u0131d\u0131r.<\/strong> (Bu durum, ki\u015finin bunu fark etmek zorunda oldu\u011fu anlam\u0131na gelmez, deneyimin de\u011fi\u015fmesi her zaman de\u011fi\u015fimi deneyimlemeyi gerektirmez (Simons & Levin, 1997)).<\/p>\n\n\n\n

Block taraf\u0131ndan da an\u0131lan di\u011fer ele\u015ftiriler ikna edicilikle ilgili kayg\u0131lara i\u015faret eder. Bechtel ve Mundale\u2019\u0131n<\/strong> (1999) erken d\u00f6nem arg\u00fcmanlar\u0131n\u0131 temel alan Godfrey-Smith, canl\u0131 par\u00e7alar\u0131n (\u00f6rne\u011fin n\u00f6ronlar\u0131n) cans\u0131z par\u00e7alarla (\u00f6rne\u011fin silikon bile\u015fenlerle) ikamesi ilerledik\u00e7e, sistemin i\u00e7 i\u015fleyi\u015finde ve toplam davran\u0131\u015f\u0131nda ka\u00e7\u0131n\u0131lmaz olarak farkl\u0131l\u0131klar ortaya \u00e7\u0131kaca\u011f\u0131n\u0131 savunur<\/strong> (Godfrey-Smith, 2016). Bu farkl\u0131l\u0131klar, metabolik k\u0131s\u0131tlar\u0131n, elektromanyetik alanlar\u0131n, ince taneli zamanlama ili\u015fkilerinin ve muhtemelen ba\u015fka etkenlerin n\u00f6ral faaliyet \u00fczerindeki etkileri nedeniyle ortaya \u00e7\u0131kar (bunun, daha \u00f6nce verilen i\u015flemlemeselci olmayan i\u015flevsellik \u00f6rnekleriyle benzerli\u011fine dikkat ediniz). Sistem hem i\u00e7ten hem d\u0131\u015ftan farkl\u0131 davran\u0131yorsa, bilincin mutlaka korunaca\u011f\u0131 varsay\u0131m\u0131 i\u00e7in bask\u0131 da ortadan kalkar.<\/strong><\/p>\n\n\n\n

Rosa Cao bu arg\u00fcman hatt\u0131n\u0131 geni\u015fletir ve zihinsel i\u015flevlerin birbirleriyle ve alt\u0131nda yatan fizyolojik ve metabolik temelleriyle nas\u0131l s\u0131k\u0131ca i\u00e7 i\u00e7e ge\u00e7ti\u011fini ayr\u0131nt\u0131land\u0131ran pek \u00e7ok betimleme sunar (Cao, 2022). Ona g\u00f6re zihinsel i\u015flevler evrim yoluyla ortaya \u00e7\u0131kt\u0131\u011f\u0131 i\u00e7in, bu i\u015flevlerin ger\u00e7ekleyicileri zaman i\u00e7inde bir\u00e7ok ba\u011f\u0131ml\u0131l\u0131k ve ko\u015fulluluk [contingencies<\/em>] biriktirmi\u015ftir. Bunlar\u0131n baz\u0131lar\u0131 keyf\u00ee g\u00f6r\u00fcnebilir ve t\u00fcm\u00fc birlikte, beynin i\u00e7 \u00f6rg\u00fctlenmesinde Wimsatt\u2019\u0131n<\/strong> (1986) \u201c\u00fcretken yerle\u015fiklik\u201d [generative entrenchment<\/em>] dedi\u011fi olguya yol a\u00e7ar.<\/strong> S\u00f6z konusu \u00fcretken yerle\u015fiklik, muhtemel alternatif uygulamalara [implementation<\/em>] pek \u00e7ok k\u0131s\u0131t dayat\u0131r ve bu k\u0131s\u0131tlar b\u00fcy\u00fck ihtimalle silikon gibi alternatif substratlar\u0131 d\u0131\u015far\u0131da b\u0131rak\u0131r.<\/strong><\/p>\n\n\n\n

Buna pek \u00e7ok \u00f6rnek verilebilir. Cao\u2019nun (2022) sundu\u011fu \u00f6rneklerden biri, bir\u00e7ok n\u00f6ral i\u015flevde g\u00f6r\u00fclen bir n\u00f6roiletken olan nitrik oksitle ilgilidir. Nitrik oksit, h\u00fccre zarlar\u0131ndan serbest\u00e7e yay\u0131lmas\u0131 hasebiyle ay\u0131rt edicidir. Bu nedenle herhangi bir silikon ikamenin, bu maddenin (ya da sim\u00fcle edilmi\u015f bir kar\u015f\u0131l\u0131\u011f\u0131n\u0131n) saptamas\u0131 gereken (belki de sim\u00fcle edilmi\u015f) konumlar\u0131, fiilen imk\u00e2ns\u0131z say\u0131dad\u0131r.[11]<\/a> Bir ba\u015fka \u00f6rnek, baz\u0131 n\u00f6ronal aksiyon potansiyellerinin [spike<\/em>] zararl\u0131 reaktif oksijen t\u00fcrlerinin birikimine kar\u015f\u0131 n\u00f6ronlar\u0131 koruyarak homeostatik bir rol oynad\u0131\u011f\u0131n\u0131 ortaya koyan yeni bir \u00e7al\u0131\u015fmad\u0131r (Chintaluri & Vogels, 2023). Bu bulgu, bir n\u00f6ronun i\u015flevsel rol\u00fc a\u00e7\u0131s\u0131ndan kilit \u00f6nemde say\u0131lan ate\u015fleme etkinli\u011finin metabolik temellerinden ayr\u0131\u015ft\u0131r\u0131lamayaca\u011f\u0131 anlam\u0131na gelir. <\/strong>Hatta bakterilerdeki kemotaktik davran\u0131\u015f bile metabolik s\u00fcre\u00e7lerden ayr\u0131 d\u00fc\u015f\u00fcn\u00fclemeyebilir (Alexandre, 2010; Egbert ve di\u011ferleri, 2010). Bu durum, metabolizma ile zihin aras\u0131ndaki ileride d\u00f6nece\u011fim bir ba\u011fa i\u015faret eder.<\/p>\n\n\n\n

Derek Shiller daha soyut d\u00fczeyde kayg\u0131lar dile getirir (Shiller, 2024). Ona g\u00f6re i\u015flevselcili\u011fin serbest bir yorumu, \u015feyleri belirli bi\u00e7imlerde par\u00e7alara ay\u0131rarak her t\u00fcrden \u015feye zihinsel durum ve bilin\u00e7 atfedebilir. Uygun bir bi\u00e7imde \u201cpar\u00e7alara ayr\u0131ld\u0131\u011f\u0131\u201d [carved up<\/em>] takdirde bir kova su, k\u0131sa bir s\u00fcre i\u00e7in bir beynin i\u015flevsel \u00f6rg\u00fctlenmesini yans\u0131tabilir.<\/strong>[12]<\/a> Bunu \u00f6nlemek i\u00e7in neyin bir sistemin par\u00e7alar\u0131 say\u0131labilece\u011fine k\u0131s\u0131tlar getirmek gerekir.<\/strong> Shiller, par\u00e7alar\u0131n i\u015flevsel rollerini i\u00e7kin nedensel g\u00fc\u00e7leri sayesinde oynamalar\u0131 gerekmesi gibi \u00e7e\u015fitli k\u0131s\u0131tlar belirler. Bu k\u0131s\u0131tlar\u0131n mevcut sinir a\u011f\u0131 modelleri taraf\u0131ndan kar\u015f\u0131lanmad\u0131\u011f\u0131n\u0131 ve genel olarak biyolojik olmayan substratlar i\u00e7in kar\u015f\u0131lanmas\u0131n\u0131n g\u00fc\u00e7 olabilece\u011fini savunur.<\/strong><\/p>\n\n\n\n

Burada ele ald\u0131\u011f\u0131m\u0131z ikna edicilik arg\u00fcmanlar\u0131, silikon substrat esnekli\u011fini \u00e7\u00fcr\u00fctmez, ancak bunun mutlaka ge\u00e7erli oldu\u011fu varsay\u0131m\u0131n\u0131 sarsar.<\/strong>[13]<\/a> Beyinler, yapt\u0131klar\u0131<\/em> ile ne olduklar\u0131n\u0131<\/em> birbirinden ay\u0131rman\u0131n \u00e7ok zor, belki de imk\u00e2ns\u0131z oldu\u011fu t\u00fcrden \u015feyler gibi g\u00f6r\u00fcn\u00fcr.<\/strong>[14]<\/a><\/p>\n\n\n\n

<\/div>\n\n\n\n

3.4 \u0130\u015flemlemenin daha geni\u015f bi\u00e7imleri<\/strong><\/p>\n\n\n\n

<\/div>\n\n\n\n

Substrat esnekli\u011fi, Turing d\u00fcnyas\u0131n\u0131n baz\u0131 k\u0131s\u0131tlar\u0131n\u0131 gev\u015feten daha geni\u015f i\u015flemleme bi\u00e7imleri taraf\u0131ndan da sorgulan\u0131r. Daha \u00f6nce belirtildi\u011fi \u00fczere, neyin \u201ci\u015flemleme\u201d say\u0131laca\u011f\u0131 konusunda kayda de\u011fer bir tart\u0131\u015fma vard\u0131r <\/strong>(N. G. Anderson & G. Piccinini, 2024; Cantwell Smith, 2002; Chalmers, 1994, 1996b; Kirkpatrick, 2022; Searle, 1990).<\/strong> Burada ayr\u0131nt\u0131ya inemesem de birka\u00e7 \u00f6rne\u011fi vurgulamak faydal\u0131 olacakt\u0131r.<\/p>\n\n\n\n

Yak\u0131n zamanda \u00f6nerilen \u00f6l\u00fcml\u00fc i\u015flemleme<\/em> [mortal computation<\/em>] kavram\u0131 \u00f6zellikle dikkat \u00e7ekicidir (Hinton, 2022; Ororbia & Friston, 2023). Standart Turing i\u015flemlemesi \u201c\u00f6l\u00fcms\u00fcz\u201dd\u00fcr. Varl\u0131\u011f\u0131 ve yararl\u0131l\u0131\u011f\u0131, herhangi bir donan\u0131m \u00f6rne\u011finin \u00f6mr\u00fcn\u00fc a\u015far. Bu durum, yaz\u0131l\u0131m\u0131n ilkece ve uygulamada donan\u0131mdan ayr\u0131\u015ft\u0131r\u0131labilir olmas\u0131 gerekti\u011fini s\u00f6yleyen temel bilgisayar bilimi ilkesini yans\u0131t\u0131r, b\u00f6ylece ayn\u0131 algoritma farkl\u0131 donan\u0131mlar \u00fczerinde y\u00fcr\u00fct\u00fcld\u00fc\u011f\u00fcnde ayn\u0131 sonucu verecektir. Fakat \u00f6l\u00fcms\u00fcz i\u015flemleme pahal\u0131d\u0131r. 1\u2019lerin 1 (ve 0\u2019lar\u0131n 0) olarak kalmas\u0131n\u0131 temin etmek i\u00e7in s\u00fcrekli hata d\u00fczeltimi yap\u0131lmas\u0131 gerektirir. Algoritmalar ve modeller karma\u015f\u0131kla\u015ft\u0131k\u00e7a hata d\u00fczeltiminin, dolay\u0131s\u0131yla i\u015flemlemesel \u00f6l\u00fcms\u00fczl\u00fc\u011f\u00fcn i\u015flemleme ve enerji maliyetleri h\u0131zla artar.<\/strong><\/p>\n\n\n\n

Bu arg\u00fcman\u0131n bir sonucu \u015fudur ki y\u00fcksek enerji verimlili\u011fine sahip biyolojik beyinler, \u00f6l\u00fcms\u00fcz i\u015flemlemeler ger\u00e7ekle\u015ftiriyor olamaz. E\u011fer bir t\u00fcr i\u015flemleme ger\u00e7ekle\u015ftiriyorlarsa, bunlar b\u00fcy\u00fck olas\u0131l\u0131kla \u00f6l\u00fcml\u00fc<\/em> i\u015flemlemelerdir, bu da i\u015flemlemelerin onlar\u0131 ger\u00e7ekle\u015ftiren \u201cdonan\u0131m\u201ddan (ya da \u201c\u0131slak donan\u0131m\u201ddan [wetware<\/em>]) ayr\u0131lamayaca\u011f\u0131 anlam\u0131na gelir.<\/strong> Bu durum, s\u00f6z konusu (\u00f6l\u00fcml\u00fc) i\u015flemlemelerin \u00e7oklu ger\u00e7ekle\u015febilirli\u011fine ve substrat esnekli\u011fine k\u0131s\u0131tlar getirir. \u00d6zellikle de bilin\u00e7li YZ i\u00e7in gereken substrat esnekli\u011fi muhtemelen ge\u00e7erli de\u011fildir. \u00c7\u00fcnk\u00fc (geleneksel) YZ, i\u015flemlemesel \u00f6l\u00fcms\u00fczl\u00fc\u011f\u00fc varsayan bir uygulama paradigmas\u0131na dayan\u0131r.<\/strong> \u0130\u015flemlemeselci bir zihin g\u00f6r\u00fc\u015f\u00fcnden kaynaklanan s\u0131n\u0131rlamalara dayand\u0131\u011f\u0131 i\u00e7in, bunu bilin\u00e7li YZ\u2019nin ikna edicili\u011fine kar\u015f\u0131 k\u0131\u015fk\u0131rt\u0131c\u0131 bir arg\u00fcman olarak g\u00f6r\u00fcyorum.[15]<\/a><\/p>\n\n\n\n

Gualtiero Piccinini\u2019nin ileri s\u00fcrd\u00fc\u011f\u00fc ili\u015fkili bir g\u00f6r\u00fc\u015fe g\u00f6re, biyolojik i\u015flevler n\u00f6ral i\u015flemleme<\/em>yi i\u00e7erir. Piccinini\u2019nin tan\u0131m\u0131nda bu, hem s\u00fcrekli hem de kesintili [discrete<\/em>] sinyalle\u015fmeyi kapsar ve bu nedenle standart yapay sinir a\u011flar\u0131n\u0131n ger\u00e7ekle\u015ftirdi\u011fi \u201cn\u00f6ral i\u015flemlemelerden\u201d farkl\u0131d\u0131r, onlara k\u0131yasla substrata daha \u00e7ok ba\u011f\u0131ml\u0131d\u0131r<\/strong> (Piccinini, 2020, bl. 13).<\/strong> Benzer \u015fekilde Kay Kirkpatrick\u2019in \u00f6nerdi\u011fi biyolojik i\u015flemleme<\/em> g\u00f6r\u00fc\u015f\u00fc de canl\u0131 sistemlerin hangi ko\u015fullarda \u201ci\u015flemleme yap\u0131yor\u201d denebilece\u011fine dair belli ba\u015fl\u0131 ko\u015fullar ortaya koyar (Kirkpatrick, 2022).<\/p>\n\n\n\n

\u00d6te yandan, standart bilgisayar bilimi paradigmas\u0131na teknolojik alternatifler uzun zamand\u0131r mevcuttur. Tarihi en az\u0131ndan Antik Yunan\u2019a kadar uzanan[16]<\/a> analog i\u015flemleme<\/em>, s\u00fcreklilik arz eden de\u011fi\u015fkenler i\u00e7erir ve tipik olarak (y\u00fckselte\u00e7ler ve kapasit\u00f6rler gibi) altta yatan d\u00fczeneklerin \u00f6zg\u00fcl \u00f6zelliklerine ba\u011f\u0131ml\u0131d\u0131r (Ullman, 2022). Sinirbi\u00e7imsel <\/em>[neuromorphic] i\u015flemleme<\/em>, modern CPU ve GPU\u2019larla kar\u015f\u0131la\u015ft\u0131r\u0131ld\u0131\u011f\u0131nda, donan\u0131m d\u00fczeyinde biyolojik sinir sistemleriyle daha fazla ortakl\u0131k ta\u015f\u0131r (Mead, 2020; Schuman ve di\u011ferleri, 2022). Baz\u0131 sinirbi\u00e7imsel yongalar (\u00e7ipler), donan\u0131m \u00f6\u011feleri aras\u0131nda ileti\u015fim i\u00e7in aksiyon potansiyelleri [spike<\/em>] kullan\u0131r ve ba\u015fka yararlar\u0131n yan\u0131 s\u0131ra daha y\u00fcksek enerji verimlili\u011fi sa\u011flar (Hochstetter ve di\u011ferleri, 2021; Merolla ve di\u011ferleri, 2014). Di\u011fer sinirbi\u00e7imsel yakla\u015f\u0131mlar analog i\u015flemlemeden yararlan\u0131r, b\u00f6ylece i\u015flev, substrata daha do\u011frudan ba\u011flan\u0131r (Douglas ve di\u011ferleri, 1995). Baz\u0131 yeni yakla\u015f\u0131mlar ise, standart i\u015flemlemesel girdi\/\u00e7\u0131kt\u0131 ayg\u0131tlar\u0131yla aray\u00fczlenmi\u015f biyolojik sinir a\u011flar\u0131n\u0131 kullan\u0131r (Morales Pantoja ve di\u011ferleri, 2023).[17]<\/a> Donan\u0131m analog veya sinirbi\u00e7imsel hale geldik\u00e7e, ger\u00e7ekle\u015ftirdi\u011fi i\u015flevler ya da i\u015flemlemeler bak\u0131m\u0131ndan substrat esnekli\u011finin azalmas\u0131 makuld\u00fcr.<\/strong>[18]<\/a><\/p>\n\n\n\n

Bu daha geni\u015f i\u015flemleme anlay\u0131\u015flar\u0131, genel olarak i\u015flemlemeselci bir bak\u0131\u015f\u0131n i\u00e7inden substrat esnekli\u011fine itiraz eder ve bunu hem biyolojik hem de yapay \u00fcretim y\u00f6n\u00fcnden yapar.<\/strong> Standart Turing i\u015flemlemesinin, biyolojik beyinlerde bilincin temelinde yer alma ihtimali bulunan \u00f6l\u00fcml\u00fc (ya da ba\u015fka t\u00fcr) i\u015flemlemeleri yine de sim\u00fcle edebilece\u011fi ileri s\u00fcr\u00fclebilir. Ancak i\u015flemlemelerin kendisi farkl\u0131 olacakt\u0131r ve sim\u00fclasyonlar\u0131n taml\u0131\u011f\u0131n\u0131 [exact<\/em>] garanti etmek m\u00fcmk\u00fcn de\u011fildir. Ayr\u0131ca, s\u00f6z konusu i\u015flemleme anlay\u0131\u015flar\u0131ndan herhangi biri i\u00e7in i\u015flemlemeselci i\u015flevselcili\u011fi yeniden tan\u0131mlamak da m\u00fcmk\u00fcnd\u00fcr. Fakat substrat esnekli\u011fi azald\u0131\u011f\u0131 i\u00e7in bu yakla\u015f\u0131m\u0131n dayand\u0131\u011f\u0131 motivasyon da zay\u0131flam\u0131\u015f olur. Uygulamada, Turing d\u00fcnyas\u0131n\u0131n sundu\u011funa benzer \u015fekilde, bilin\u00e7le ili\u015fkili (\u00f6l\u00fcml\u00fc, sinirsel vb.) i\u015flemlemelerin (sinirbi\u00e7imsel, analog vb.) donan\u0131mda ger\u00e7ekle\u015ftirilebilece\u011fi g\u00fcvencesi verilemez.<\/strong><\/p>\n\n\n\n

\u201c\u0130\u015flemlemesel\u201d olan\u0131n s\u0131n\u0131rlar\u0131n\u0131 de\u011fi\u015ftirmek, \u201ci\u015flemlemesel olmayan\u201d\u0131n s\u0131n\u0131rlar\u0131n\u0131 da de\u011fi\u015ftirir. Daha geni\u015f i\u015flemleme anlay\u0131\u015flar\u0131 aras\u0131ndan se\u00e7im yapma imk\u00e2n\u0131 olsa bile, i\u015flemlemesel olmayan i\u015flevler bili\u015f ve bilin\u00e7 i\u00e7in yine de kritik olabilir ve beyinlerin herhangi bir \u015fekilde \u201ci\u015flemleme\u201d yap\u0131p yapmad\u0131\u011f\u0131 sorusu a\u00e7\u0131k kal\u0131r.<\/strong><\/p>\n\n\n\n

<\/div>\n\n\n\n

3.5 Belirim ve \u00f6l\u00e7eklerin ayr\u0131m\u0131<\/em><\/p>\n\n\n\n

<\/div>\n\n\n\n

Substrat esnekli\u011fi \u00fczerine ba\u015fka yararl\u0131 bir bak\u0131\u015f, fizikteki hiyerar\u015fik belirim [emergence<\/em>] olgusundan gelir. Fizik\u00e7iler, \u201c\u00f6l\u00e7eklerin ayr\u0131m\u0131\u201dndan [separation of scales<\/em>] s\u00f6z eder, buna g\u00f6re daha y\u00fcksek, yani belirmi\u015f [emergent<\/em>] betimleme d\u00fczeylerinde yasalar form\u00fcle edilebilir ve bu, daha ince ayr\u0131nt\u0131lar\u0131 bilmeyi gerektirmez. M\u00fchendisler, kuantum mekani\u011fini bilmeden k\u00f6pr\u00fc in\u015fa edebilir. Bilgisayar programc\u0131lar\u0131, uygulaman\u0131n ayr\u0131nt\u0131lar\u0131yla ilgilenmeden yaz\u0131l\u0131m yazabilir. Yaz\u0131l\u0131m\u0131n substrat esnekli\u011fi (ve \u00f6l\u00fcms\u00fczl\u00fc\u011f\u00fc), \u00f6l\u00e7eklerin bu g\u00fc\u00e7l\u00fc ayr\u0131m\u0131na dayan\u0131r.<\/p>\n\n\n\n

Yak\u0131n zamanda karma\u015f\u0131k sistemlerde \u00f6l\u00e7eklerin ayr\u0131m\u0131n\u0131 temellendiren \u201cenformasyonel kapanma\u201d ve \u201cnedensel kapanma\u201d [closure<\/em>] fikirlerini bi\u00e7imselle\u015ftiren ampirik \u00f6l\u00e7\u00fcler geli\u015ftirildi (\u00f6r. Barnett & Seth, 2023; Rosas ve di\u011ferleri, 2024). Makroskopik (daha \u00fcst d\u00fczeyde) bir de\u011fi\u015fken ya da \u201ciri tanelilik\u201d [coarse graining<\/em>], yap\u0131lan m\u00fcdahalelerin \u00fcst d\u00fczey sonu\u00e7lar\u0131 belirlemeye yetti\u011fi durumda nedensel olarak kapal\u0131, bu de\u011fi\u015fkenin g\u00f6zlemlerinin \u00fcst d\u00fczey sonu\u00e7lar\u0131 \u00f6ng\u00f6rmeye yetti\u011fi durumda enformasyonel olarak kapal\u0131 say\u0131l\u0131r. Elbette bu sonu\u00e7lar belirlenebilir ya da \u00f6ng\u00f6r\u00fclebilir oldu\u011fu \u00f6l\u00e7\u00fcde. Ba\u015fka bir ifadeyle daha alt, mikroskobik d\u00fczeylere ili\u015fkin bilgi (ya da o d\u00fczeylerdeki m\u00fcdahaleler), makro d\u00fczeyde yap\u0131labileceklerin \u00f6tesinde bir \u00f6ng\u00f6r\u00fc g\u00fcc\u00fc veya sonu\u00e7lar\u0131 belirleme imk\u00e2n\u0131 sa\u011flamaz. Bu durum (d\u00fczg\u00fcn i\u015fleyen geleneksel) bilgisayarlarda tasar\u0131m gere\u011fi g\u00fcvence alt\u0131ndad\u0131r. Ancak bu kapanma bi\u00e7imlerinin karma\u015f\u0131k sistemler i\u00e7in genel ge\u00e7er oldu\u011funu varsaymak hatal\u0131 olur. \u00d6zellikle evrimsel tasar\u0131m\u0131n do\u011fas\u0131nda bulunan \u00fcretken yerle\u015fiklik sayesinde biyolojik sistemlerde bu t\u00fcr kapanmalar k\u0131smen ya da tamamen g\u00f6r\u00fclmeyebilir.<\/strong><\/p>\n\n\n\n

Nedensel ve enformasyonel kapanmay\u0131 \u00f6l\u00e7me yetene\u011fi, substrat esnekli\u011finin verili kabul edilemeyece\u011fini vurgular.<\/strong> Uygulamada, bu t\u00fcr \u00f6l\u00e7\u00fcler, sinirsel dinamiklerin, betimlemenin daha ince \u00f6l\u00e7ekli d\u00fczeylerinden ayr\u0131\u015ft\u0131r\u0131l\u0131p ayr\u0131\u015ft\u0131r\u0131lamayaca\u011f\u0131n\u0131 ve ayr\u0131\u015ft\u0131r\u0131labilecekse bunun hangi \u00f6l\u00e7\u00fcde m\u00fcmk\u00fcn oldu\u011funu ampirik olarak belirlemeye yard\u0131mc\u0131 olabilir.<\/p>\n\n\n\n

<\/div>\n\n\n\n

3.6 \u0130\u015flemlemeselci i\u015flevselcili\u011fin \u00f6tesinde<\/em><\/p>\n\n\n\n

<\/div>\n\n\n\n

Buraya kadar ileri s\u00fcr\u00fclen arg\u00fcmanlar, i\u015flemlemeselci i\u015flevselcili\u011fi \u00e7\u00fcr\u00fctmez. Ancak n\u00f6robiyolojik i\u015flevlerin muhtemelen s\u0131n\u0131rl\u0131 substrat esnekli\u011fine sahip oldu\u011funu ve benzer k\u0131s\u0131tlar\u0131 olan daha geni\u015f i\u015flemleme bi\u00e7imlerinin de varl\u0131\u011f\u0131n\u0131 g\u00f6stererek bu g\u00f6r\u00fc\u015f\u00fcn ikna edicili\u011fini ve cazibesini azalt\u0131r.<\/strong> Burada yaln\u0131zca ge\u00e7erken de\u011finebilece\u011fim bir\u00e7ok ba\u015fka arg\u00fcman da ileri s\u00fcr\u00fclebilir. Bunlar aras\u0131nda s\u00f6z dizimi ile anlam bilgisi aras\u0131ndaki ayr\u0131m (Cole, 2023; Searle, 1980) yer al\u0131r. Bu ayr\u0131m, bilincin g\u00f6zlemciye ba\u011f\u0131ml\u0131 olabilece\u011fini ima eder.<\/strong>[19]<\/a> Y\u00f6nelimsellik \u00fczerine felsefi de\u011ferlendirmeler (Shagrir, 2010) ve Roger Penrose\u2019un G\u00f6del teoreminden hareketle geli\u015ftirdi\u011fi arg\u00fcman (Penrose, 1989) da bu ba\u011flamdad\u0131r.<\/p>\n\n\n\n

T\u00fcm bunlar \u0131\u015f\u0131\u011f\u0131nda, i\u015flemlemeselci i\u015flevselcili\u011fin ve yayg\u0131n substrat esnekli\u011finin do\u011fru olmas\u0131 gerekti\u011fini basit\u00e7e varsaymamal\u0131y\u0131z. Bilin\u00e7li YZ, i\u015flemlemeselci i\u015flevselcili\u011fin do\u011frulu\u011funa ve substrat esnekli\u011finin en az\u0131ndan silikon i\u00e7in ge\u00e7erli olmas\u0131na ba\u011fl\u0131d\u0131r. \u00d6yleyse silikon substratta \u00e7al\u0131\u015fan dijital i\u015flemlemeler olarak anlad\u0131\u011f\u0131m\u0131z bilin\u00e7li YZ\u2019nin m\u00fcmk\u00fcn olup olmad\u0131\u011f\u0131n\u0131 sorgulamak i\u00e7in makul gerek\u00e7elerimiz vard\u0131r.<\/strong><\/p>\n\n\n\n

Beynin nas\u0131l bir \u015fey oldu\u011funa bak\u0131nca, i\u015flemlemeselci i\u015flevselcili\u011fin bu denli yayg\u0131n kabul g\u00f6rm\u00fc\u015f olmas\u0131 belki de \u015fa\u015f\u0131rt\u0131c\u0131d\u0131r. Beynin i\u00e7ine bakt\u0131\u011f\u0131m\u0131zda, bilgisayarlarda donan\u0131m ile yaz\u0131l\u0131m aras\u0131nda buldu\u011fumuz t\u00fcrden keskin bir \u201czihin yaz\u0131l\u0131m\u0131\u201d [mindware<\/em>] \/ \u201c\u0131slak donan\u0131m\u201d [wetware<\/em>] ayr\u0131m\u0131 g\u00f6rmeyiz.<\/strong><\/p>\n\n\n\n

Beyin etkinli\u011fi \u00f6r\u00fcnt\u00fcleri uzay ve zamanda \u00e7oklu \u00f6l\u00e7eklerde evrilir, s\u00fcrekli olarak n\u00f6roiletkenlerin kimyasal yay\u0131l\u0131m\u0131n\u0131, sinir a\u011flar\u0131n\u0131n kendilerinin fiziksel yap\u0131s\u0131n\u0131 (sinaptik plastisite ve sinaptogenez) ve h\u00fccre altyap\u0131s\u0131n\u0131n \u00e7e\u015fitli \u00f6zelliklerini etkilerler ve bunlardan etkilenirler. Hepsi metabolik etkinli\u011fin molek\u00fcler f\u0131rt\u0131nas\u0131yla derinden i\u00e7 i\u00e7e ge\u00e7mi\u015ftir ya da \u00fcretken bir yerle\u015fiklik g\u00f6sterir. Tek bir n\u00f6ron ba\u015fl\u0131 ba\u015f\u0131na karma\u015f\u0131k bir biyolojik sistemdir. B\u00fct\u00fcnl\u00fc\u011f\u00fcn\u00fc korumak ve maddi varl\u0131\u011f\u0131n\u0131n ko\u015fullar\u0131n\u0131 yeniden \u00fcretmek i\u00e7in durmaks\u0131z\u0131n \u00e7al\u0131\u015f\u0131r. Bu durumda beyni bilgisayara benzeten metafor giderek daha elveri\u015fsiz bir kurgu haline gelmektedir. T\u00fcm bu \u00e7ok \u00f6l\u00e7ekli biyolojik etkinli\u011fi, yaln\u0131zca do\u011fru t\u00fcrden bir i\u015flemlemeye imk\u00e2n sa\u011flayan bir arka plan olarak g\u00f6rmek dar g\u00f6r\u00fc\u015fl\u00fc g\u00f6r\u00fcnmektedir.<\/strong><\/p>\n\n\n\n

“Beyni bilgisayara benzeten metafor giderek daha elveri\u015fsiz bir kurgu haline gelmektedir, t\u00fcm bu \u00e7ok \u00f6l\u00e7ekli biyolojik etkinli\u011fi, yaln\u0131zca do\u011fru t\u00fcrden bir i\u015flemlemeye imk\u00e2n sa\u011flayan bir arka plan olarak g\u00f6rmek dar g\u00f6r\u00fc\u015fl\u00fc g\u00f6r\u00fcnmektedir.”<\/p><\/blockquote><\/figure>\n\n\n\n

Ba\u015fka olas\u0131l\u0131klar da vard\u0131r. Bilin\u00e7, alternatif i\u015flemleme bi\u00e7imlerine dayan\u0131yor olabilir. Ancak, belirtildi\u011fi gibi, bu alternatiflerin Turing i\u015flemlemesinin sundu\u011fu substrat esnekli\u011fine sahip olmamalar\u0131 nedeniyle biyolojik olmayan alternatif substratlarda uygulanamayabilirler. Bu durumda en ba\u015fta i\u015flemlemeselci bir g\u00f6r\u00fc\u015f\u00fc benimsemek i\u00e7in daha az motivasyon olacakt\u0131r.<\/p>\n\n\n\n

Daha geni\u015f planda, bilin\u00e7 i\u015flemlemelere de\u011fil, i\u015flevsel \u00f6rg\u00fctlenme \u00f6r\u00fcnt\u00fclerine dayan\u0131yor olabilir<\/strong> (Piccinini, 2018, 2020; Prinz, 2012). Bu i\u015flemlemeselci olmayan i\u015flevselcilik<\/em>, n\u00f6ral sistemleri i\u015flemlemeselci bir tutum varsaymadan anlamaya \u00e7al\u0131\u015fan yakla\u015f\u0131mlarla \u00f6rt\u00fc\u015f\u00fcr. S\u00f6z konusu \u00e7al\u0131\u015fmalar\u0131n \u00f6nemli b\u00f6l\u00fcm\u00fc, n\u00f6robiyoloji ve YZ \u00fczerine erken sibernetik d\u00fc\u015f\u00fcncelere uzan\u0131r, bu d\u00fc\u015f\u00fcnceler, bedenden ba\u011f\u0131ms\u0131z bir i\u015flemleme anlay\u0131\u015f\u0131ndan \u00e7ok bedenlenmi\u015flik [embodiment<\/em>], geri besleme ve denetimi vurgulam\u0131\u015ft\u0131r<\/strong> (Dupuy, 2009).<\/strong><\/p>\n\n\n\n

Bili\u015fsel bilimde yank\u0131 uyand\u0131rm\u0131\u015f \u201cdinamik sistemler\u201d yakla\u015f\u0131m\u0131 buna iyi bir \u00f6rnektir (Van Gelder, 1995). Bu yakla\u015f\u0131m, bedenlenmi\u015f ve g\u00f6m\u00fcl\u00fc [embedded<\/em>] sinirsel sistemlerin \u00f6zelliklerinin zaman i\u00e7indeki de\u011fi\u015fimini modellemek i\u00e7in genellikle diferansiyel denklemlere dayanan \u00e7e\u015fitli matematiksel ara\u00e7 ve y\u00f6ntemlerden yararlan\u0131r. Bu de\u011fi\u015fimler, hangi t\u00fcrden olursa olsun, dijital ya da analog, herhangi bir i\u015flemleme ile ili\u015fkilendirilmez (Van Gelder, 1995).[20]<\/a> Dinamik sistemler kuram\u0131 tarihsel olarak bilince odaklanmam\u0131\u015ft\u0131r, bu bak\u0131mdan i\u015flemlemeselci olmayan i\u015flevselcili\u011fi tamamlar. \u201c4E\u201d bili\u015fsel bilim ba\u015fl\u0131\u011f\u0131 alt\u0131ndaki \u00e7e\u015fitli perspektifler i\u00e7in de benzer \u015feyler s\u00f6ylenebilir, bu yakla\u015f\u0131mlar beyin-beden-d\u00fcnya etkile\u015fimlerinin bedenlenmi\u015f, g\u00f6m\u00fcl\u00fc, enaktif [enactive<\/em>] ve uzan\u0131ml\u0131 [extended<\/em>] y\u00f6nlerini \u00f6ne \u00e7\u0131kar\u0131r. <\/strong>Bu t\u00fcr \u00e7al\u0131\u015fmalarda bilince daha s\u0131k yer verilir (Cosmelli ve di\u011ferleri, 2007; Kiverstein, 2020; Stewart ve di\u011ferleri, 2010; Thompson, 2022). Bu g\u00f6r\u00fc\u015fler \u00f6nemli ayr\u0131nt\u0131larda farkl\u0131la\u015f\u0131r, fakat ortak noktalar\u0131, beynin bir bilgisayar oldu\u011fu ve zihnin \u00f6z\u00fcnde i\u015flemlemesel oldu\u011fu fikrine mesafeli durmalar\u0131d\u0131r.<\/strong><\/p>\n\n\n\n

N\u00f6robilime dayal\u0131 \u00e7e\u015fitli bilin\u00e7 kuramlar\u0131, i\u015flemlemesel de\u011fil dinamik \u00f6\u011feleri vurgular ya da bu \u00e7er\u00e7evede yeniden form\u00fcle edilebilir.<\/strong> Bunlar aras\u0131nda, bilin\u00e7li alg\u0131y\u0131 n\u00f6ral sistemlerdeki \u00e7ekici [attractor<\/em>] dinamiklerine ba\u011flayan klasik \u00e7al\u0131\u015fmalar (Freeman, 1999; Llinas ve di\u011ferleri, 1998) ile n\u00f6ral e\u015fzamanl\u0131l\u0131\u011fa, \u00f6rne\u011fin gama frekans band\u0131na vurgu yapan \u00e7al\u0131\u015fmalar yer al\u0131r (Crick & Koch, 1990). Bilincin dayand\u0131\u011f\u0131 temel n\u00f6rodinamik \u00f6zellikler olarak b\u00fct\u00fcnle\u015fme ve ayr\u0131mla\u015fmay\u0131 \u00f6ne \u00e7\u0131karan \u201cdinamik \u00e7ekirdek\u201d hipotezi de bu gruba girer. Bu hipotez daha sonra bilin\u00e7 i\u00e7in b\u00fct\u00fcnle\u015fik enformasyon kuram\u0131 [integrated information theory<\/em>] olarak yeniden g\u00fcndeme gelmi\u015ftir (Tononi ve Edelman, 1998; Albantakis ve di\u011ferleri, 2023; Tononi ve di\u011ferleri, 2016). Yinelemeli i\u015fleme [recurrent processing<\/em>] kuram\u0131 (Lamme, 2010, 2018) da ayn\u0131 \u015fekilde bu \u00e7er\u00e7evededir. \u00d6ng\u00f6r\u00fcc\u00fc i\u015fleme [predictive processing<\/em>] ve serbest enerji ilkesi de, bu kuramlar\u0131n nas\u0131l yorumland\u0131\u011f\u0131na ba\u011fl\u0131 olarak, bu kapsama girer (Clark, 2019; Friston, 2010). Bu fikirlere birazdan d\u00f6nece\u011fim. \u00c7\u00fcnk\u00fc bu yakla\u015f\u0131mlar, bilincin biyolojiye ba\u011f\u0131ml\u0131l\u0131\u011f\u0131na ili\u015fkin ortaya koyaca\u011f\u0131m sav\u0131n merkezinde yer almaktad\u0131r. Daha genel olarak n\u00f6rofenomenoloji yakla\u015f\u0131m\u0131 (\u00f6rne\u011fin Francisco Varela\u2019n\u0131n esin kayna\u011f\u0131 oldu\u011fu \u00e7izgi), sinirsel s\u00fcre\u00e7lerle fenomenolojik \u00f6zellikler aras\u0131nda k\u00f6pr\u00fc ilkeler olarak i\u015flemlemesel olmayan dinamik yap\u0131lara s\u0131k\u00e7a ba\u015fvurur <\/strong>(Thompson & Varela, 2001; Varela, 1996).<\/strong><\/p>\n\n\n\n

Dinamik n\u00f6robilimsel ve (n\u00f6ro)fenomenolojik yakla\u015f\u0131mlar umut vericidir. Bununla birlikte, belirli bir kuram\u0131n \u00f6nerdi\u011fi dinamik ilkelerin alternatif bir maddi yap\u0131da uygulanmas\u0131n\u0131n bilinci ger\u00e7ekle\u015ftirmeye yeterli olup olmayaca\u011f\u0131 konusunda genellikle ba\u011flay\u0131c\u0131 bir iddia ortaya koymazlar.<\/strong> (Burada \u201cbilin\u00e7li YZ\u201d ifadesini kullanmaktan \u00f6zellikle imtina ediyorum, \u00e7\u00fcnk\u00fc art\u0131k ne i\u015flemlemelerin uygulanmas\u0131ndan s\u00f6z ediyoruz ne de zek\u00e2dan.) Ayr\u0131ca, daha geni\u015f i\u015flemleme anlay\u0131\u015flar\u0131nda oldu\u011fu gibi, i\u015flemlemesel olmayan dinamik s\u00fcre\u00e7leri alternatif substratlarda uygulamak da substrat esnekli\u011fi sorununu g\u00fcndeme getirir, \u00e7\u00fcnk\u00fc standart i\u015flemlemeselcili\u011fin ilkece sa\u011flad\u0131\u011f\u0131 substrat ba\u011f\u0131ms\u0131zl\u0131\u011f\u0131na art\u0131k ba\u015fvurulamaz.<\/strong><\/p>\n\n\n\n

<\/div>\n\n\n\n

3.7 Sim\u00fclasyon, uygulama ve ger\u00e7ekle\u015ftirme<\/em><\/p>\n\n\n\n

<\/div>\n\n\n\n

Bu noktada, \u015fimdiye dek olduk\u00e7a \u00f6rt\u00fck b\u0131rak\u0131lm\u0131\u015f olan sim\u00fclasyon, uygulama ve ger\u00e7ekle\u015ftirme aras\u0131ndaki ili\u015fkileri netle\u015ftirmekte yarar var. N\u00f6robiyolojik bir s\u00fcre\u00e7 gibi hedef bir d\u00fczenek ya da sistem s\u00f6z konusu oldu\u011funda bu terimleri a\u015fa\u011f\u0131daki gibi kullan\u0131yorum. D\u00fczenek, bir (Turing) i\u015flemlemesi olarak soyutlan\u0131r ve ard\u0131ndan bu i\u015flemleme dijital bir bilgisayarda uygulan\u0131rsa sim\u00fcle edilmi\u015f<\/em> olur. Her d\u00fczenek sim\u00fcle edilebilir; fakat bu sim\u00fclasyonlar yaln\u0131zca yakla\u015f\u0131k olabilir ve hatta \u00e7ok b\u00fcy\u00fck hatalar i\u00e7erebilir.<\/strong> Alternatif olarak, hedef d\u00fczene\u011fin daha ince taneli ya da ba\u015fka i\u015flemlemesel olmayan y\u00f6nleri ba\u015fka yollarla (\u00f6rne\u011fin sinirbi\u00e7imsel m\u00fchendislik ya da sentetik biyoloji arac\u0131l\u0131\u011f\u0131yla) uygulanabilir<\/em>. Bu yakla\u015f\u0131m\u0131n s\u0131n\u0131r\u0131, fiziksel olarak \u00f6zde\u015f bir t\u0131pk\u0131 \u00fcretimdir [replication<\/em>]. Son olarak \u015fu soruyu sorabiliriz: Bir d\u00fczene\u011fin sim\u00fclasyonu ya da o d\u00fczene\u011fin ba\u015fka t\u00fcrden (\u00f6rne\u011fin i\u015flemlemesel olmayan) uygulamalar\u0131, bilin\u00e7 gibi bir ba\u015fka \u00f6zelli\u011fi ger\u00e7ekle\u015ftirecek<\/em> (\u00f6rnekleyecek) midir?<\/strong><\/p>\n\n\n\n

Sim\u00fclasyon ile uygulama aras\u0131ndaki ayr\u0131m\u0131n keskin olmas\u0131 gerekmez. Bir sistemi daha geni\u015f i\u015flemleme bi\u00e7imlerini kullanarak da \u201csim\u00fcle\u201d edebiliriz. \u00d6rne\u011fin analog i\u015flemleme yoluyla ya da fiziksel nedensel modeller kullanarak. Daha ince taneli, nedensel ve muhtemelen i\u015flemlemesel olmayan \u00f6zelliklerin d\u00e2hil edilmesine izin verildi\u011finde bu \u201csim\u00fclasyonlar\u201d, \u201cuygulamalar\u201da yakla\u015facakt\u0131r.<\/p>\n\n\n\n

Genel olarak, bir \u015feyi sim\u00fcle etmek (\u00f6rne\u011fin bir ya\u011fmur f\u0131rt\u0131nas\u0131n\u0131 ya da sindirim s\u00fcrecini) o \u015feyi ger\u00e7ekle\u015ftirmekle (\u00f6rneklemekle) e\u015fde\u011fer de\u011fildir. Sim\u00fclasyonlar, sim\u00fcle edilen \u015feylerin nedensel g\u00fc\u00e7lerinden ve i\u00e7kin \u00f6zelliklerinden yoksundur. X\u2019in sim\u00fclasyonunun X\u2019i kesin olarak ger\u00e7ekle\u015ftirmesi i\u00e7in, i\u015flemlemenin X i\u00e7in yeterli oldu\u011fundan emin olmak gerekir.<\/strong> X\u2019in kendisi bir (Turing) i\u015flemleme s\u00fcreci oldu\u011funda bu a\u00e7\u0131k\u00e7a do\u011frudur. Ya\u011fmur f\u0131rt\u0131nalar\u0131 ve sindirim gibi kimi ba\u015fka \u015feylerde ise a\u00e7\u0131k\u00e7a yanl\u0131\u015ft\u0131r.[21]<\/a> Ya\u011fmur f\u0131rt\u0131nalar\u0131 ve sindirim en az\u0131ndan k\u0131smen i\u00e7kin maddi \u00f6zellikler (\u00f6rne\u011fin su, besin) ve nedensel g\u00fc\u00e7ler (\u00f6rne\u011fin r\u00fczg\u00e2r, metabolik tepkiler) bak\u0131m\u0131ndan tan\u0131mlan\u0131r. Hava tahmini yapan bir bilgisayarda hi\u00e7bir \u015fey \u0131slanmaz.<\/strong> Bilincin dayand\u0131\u011f\u0131 n\u00f6robiyolojik d\u00fczenekleri sim\u00fcle etmek, ister tam ister yakla\u015f\u0131k bi\u00e7imde olsun, bilinci ger\u00e7ekle\u015ftirmeye yeterli olacak m\u0131d\u0131r? Buna kesin bir \u201cevet\u201d yan\u0131t\u0131 ancak i\u015flemlemenin bilin\u00e7 i\u00e7in yeterli olmas\u0131 h\u00e2linde verilebilir. Bu da, dikkat edilirse, i\u015flemlemeselci i\u015flevselcili\u011fi varsaymay\u0131 gerektirmektedir.<\/strong>[22]<\/a><\/p>\n\n\n\n

“Genel olarak, bir \u015feyi sim\u00fcle etmek (\u00f6rne\u011fin bir ya\u011fmur f\u0131rt\u0131nas\u0131n\u0131 ya da sindirim s\u00fcrecini) o \u015feyi ger\u00e7ekle\u015ftirmekle (\u00f6rneklemekle) e\u015fde\u011fer de\u011fildir. (…) Hava tahmini yapan bir bilgisayarda hi\u00e7bir \u015fey \u0131slanmaz.<\/strong>“<\/p><\/blockquote><\/figure>\n\n\n\n

Buna kar\u015f\u0131l\u0131k, bilincin dayand\u0131\u011f\u0131 n\u00f6robiyolojik d\u00fczenekleri do\u011fru ayr\u0131nt\u0131 d\u00fczeyinde uygulamak<\/em>, tan\u0131m gere\u011fi bilinci ger\u00e7ekle\u015ftirecektir. <\/strong>Elbette bunu yapmak kolay de\u011fildir. \u0130\u015flemlemeselci modeller, i\u015flemlemesel olsun olmas\u0131n her t\u00fcrden sistem ve d\u00fczene\u011fi kolayca sim\u00fcle edebilir. Ancak bir d\u00fczene\u011fi do\u011fru ayr\u0131nt\u0131 d\u00fczeyinde uygulayabilme yetene\u011fi, o d\u00fczene\u011fin ne oldu\u011funa ve bu ayr\u0131nt\u0131 d\u00fczeyinin ne oldu\u011funa g\u00fc\u00e7l\u00fc bi\u00e7imde ba\u011fl\u0131d\u0131r.<\/p>\n\n\n\n

<\/div>\n\n\n\n

3.8 Zek\u00e2 bu tart\u0131\u015fman\u0131n neresinde?<\/strong><\/p>\n\n\n\n

<\/div>\n\n\n\n

\u015eimdiye kadar sundu\u011fum arg\u00fcmanlar yapay bilin\u00e7 a\u00e7\u0131s\u0131ndan sonu\u00e7lara odaklan\u0131yordu. Ancak beynin ne yapt\u0131\u011f\u0131<\/em>, ne oldu\u011fu<\/em>ndan ayr\u0131lam\u0131yorsa yapay zek\u00e2 a\u00e7\u0131s\u0131ndan da sonu\u00e7lar olacakt\u0131r.<\/strong> Beynin i\u015flevsel \u00f6rg\u00fctlenmesi substrat\u0131n\u0131n \u00f6zelliklerine s\u0131k\u0131ca ba\u011fl\u0131ysa, belirli zek\u00e2 t\u00fcrleri i\u00e7in gerekli olan ve yaln\u0131zca biyolojik sistemlerle s\u0131n\u0131rl\u0131 bulunan, (potansiyel olarak i\u015flemlemesel olmayan) i\u015flevsel \u00f6rg\u00fctlenme \u00f6r\u00fcnt\u00fcleri de var olabilir. <\/strong>Bahsetti\u011fimiz zek\u00e2 t\u00fcrleri ay\u0131rt edici \u015fekilde biyolojik olacakt\u0131r.<\/p>\n\n\n\n

S\u00f6z konusu i\u015flevsel \u00f6rg\u00fctlenme \u00f6r\u00fcnt\u00fcleri yeterince asl\u0131na uygun sim\u00fcle edilebilirse bilgisayarlar yine de buna kar\u015f\u0131l\u0131k gelen zek\u00e2 t\u00fcrlerini (yakla\u015f\u0131k olarak) ger\u00e7ekle\u015ftirebilir. Zek\u00e2 genel olarak yapmaya<\/em> ili\u015fkin oldu\u011fundan, yeterince isabetli sim\u00fclasyonlar, altta yatan d\u00fczenekler ve i\u00e7kin \u00f6zellikler farkl\u0131 olsa bile, ger\u00e7ekle\u015ftirme say\u0131labilir. <\/strong>\u00d6rne\u011fin b\u00fcy\u00fck dil modelleri, arka plan\u0131nda ger\u00e7ek insan dilinin temelindeki s\u00fcre\u00e7lerin zenginli\u011finden ar\u0131nd\u0131r\u0131lm\u0131\u015f bir karikat\u00fcr\u00fc bulunsa bile dilin baz\u0131 y\u00f6nlerini taklit edebilmektedir (Mitchell & Krakauer, 2023). Bilgisayarlar, oyuna dair insana benzer bir kavray\u0131\u015fa sahip olmasalar bile satran\u00e7 (en az\u0131ndan bir t\u00fcr\u00fcn\u00fc) oynarlar.[23]<\/a><\/p>\n\n\n\n

Daha k\u0131\u015fk\u0131rt\u0131c\u0131s\u0131, zek\u00e2yla ilgili baz\u0131 i\u015flevsel \u00f6rg\u00fctlenme \u00f6r\u00fcnt\u00fclerinin, silikon gibi alternatif substratlarda y\u00fcr\u00fct\u00fclen i\u015flevsel bak\u0131mdan yeterli sim\u00fclasyon ya da uygulama taraf\u0131ndan hi\u00e7bir zaman yakalanamayacak olmas\u0131d\u0131r.<\/strong> Michael Levin bu olas\u0131l\u0131\u011f\u0131 destekleyen bir sav ortaya koyar ve biyolojik malzemenin etkin do\u011fas\u0131n\u0131n, bili\u015fsel yeterliliklerin Platoncu uzay\u0131nda belli ba\u015fl\u0131 b\u00f6lgelere eri\u015fim sa\u011flad\u0131\u011f\u0131n\u0131 ileri s\u00fcrer (Levin, 2025; McMillen & Levin, 2024). Biyolojik sistemlerin bedenlenmi\u015f ve g\u00f6m\u00fcl\u00fc do\u011fas\u0131 da zek\u00e2yla ili\u015fkili i\u015flevsellik a\u00e7\u0131s\u0131ndan sim\u00fcle edilmesi g\u00fc\u00e7 veya olanaks\u0131z olabilecek k\u0131s\u0131tlar ve imk\u00e2nlar sunabilir (Bkz. B\u00f6l\u00fcm 5.7).<\/p>\n\n\n\n

Bu ba\u011flamda zek\u00e2 ile bilin\u00e7 aras\u0131ndaki fark\u0131 yeniden vurgulamak yerinde olacakt\u0131r. Zek\u00e2, en az\u0131ndan zek\u00e2n\u0131n baz\u0131 bi\u00e7imleri, girdilerin \u00e7\u0131kt\u0131larla e\u015flenmesi [mapping<\/em>] \u00fczerinden anla\u015f\u0131labilir. Bu durumda biyolojik olmakla ay\u0131rt edilen zek\u00e2 t\u00fcrleri, menzilimizin d\u0131\u015f\u0131nda kalsa ya da yaln\u0131zca gev\u015fek bi\u00e7imde sim\u00fcle edilebilse dahi i\u015flemleme yoluyla \u00f6rneklenebilir. Buna kar\u015f\u0131l\u0131k bilincin herhangi bir bi\u00e7imde<\/em> i\u015flemlemeyle \u00f6rneklenebilmesi i\u00e7in i\u015flemlemeselci i\u015flevselcili\u011fin do\u011frulu\u011funu varsaymak gerekir.<\/strong><\/p>\n\n\n\n

<\/div>\n\n\n\n

3.9 \u00d6zet<\/em><\/p>\n\n\n\n

<\/div>\n\n\n\n

Bilin\u00e7li YZ\u2019nin olabilirli\u011fi hem i\u015flemlemeselci i\u015flevselcili\u011fin ge\u00e7erli\u011fine hem de bilin\u00e7 i\u00e7in yeterli i\u015flemlemelerin YZ donan\u0131m\u0131nda (silikonda) uygulanmas\u0131n\u0131 sa\u011flayacak bir substrat esnekli\u011finin varl\u0131\u011f\u0131na ba\u011fl\u0131d\u0131r. Ancak (bilin\u00e7li) biyolojik sistemlerin ger\u00e7ekle\u015ftirdi\u011fi i\u015flevler ya da i\u015flemlemeler maddi temellerinden ayr\u0131lamayabilir.<\/strong> Bu durum, bilin\u00e7li YZ i\u00e7in gerekli substrat esnekli\u011finin do\u011fru olmayabilece\u011fi anlam\u0131na gelir. <\/strong>Di\u011fer yandan beynin i\u015flemleme yapt\u0131\u011f\u0131 ya da bilincin bir (Turing) i\u015flemleme meselesi oldu\u011fu kabullerine alternatif g\u00f6r\u00fc\u015fler de vard\u0131r. Bu g\u00f6r\u00fc\u015fler i\u015flemlemeselci i\u015flevselcili\u011fe dayanak te\u015fkil eden varsay\u0131mlar\u0131 zay\u0131flat\u0131r.<\/p>\n\n\n\n

\u00d6yleyse \u015fimdi biyolojik do\u011falc\u0131l\u0131\u011fa (nat\u00fcralizme)<\/em> ge\u00e7ebiliriz. Bu g\u00f6r\u00fc\u015fe g\u00f6re biyolojik \u00f6zellikler bilin\u00e7 i\u00e7in zorunludur.<\/strong> Bir sonraki b\u00f6l\u00fcmde, zihnin ve bilincin \u00f6zellikleriyle bunlar\u0131n biyolojik dayanaklar\u0131n\u0131n \u00f6zellikleri aras\u0131nda a\u00e7\u0131klay\u0131c\u0131 k\u00f6pr\u00fcler kurarak form\u00fcle edece\u011fim \u00e7e\u015fitli biyolojik nat\u00fcralizm t\u00fcrleri i\u00e7in gerek\u00e7eler sunaca\u011f\u0131m. Bu gerek\u00e7eler alg\u0131, bili\u015f ve bilin\u00e7 \u00fczerine belirli bir kuramsal bak\u0131\u015fa dayanmaktad\u0131r. \u00d6ng\u00f6r\u00fcc\u00fc i\u015fleme ve serbest enerji ilkesi olmak \u00fczere iki farkl\u0131 y\u00f6nden gelerek bunlar\u0131 ortak noktada bulu\u015fturaca\u011f\u0131m.<\/strong><\/p>\n\n\n\n

<\/div>\n\n\n\n

4.0 Biyolojik do\u011falc\u0131l\u0131\u011fa do\u011fru<\/strong><\/p>\n\n\n\n

<\/div>\n\n\n\n
\n

\u201cOrganizma devam etmek zorundad\u0131r, \u00e7\u00fcnk\u00fc devam etmek, onun varolu\u015funun ta kendisidir.\u201d<\/em> (Jonas, 2001)<\/p>\n<\/blockquote>\n\n\n\n

Biyolojik do\u011falc\u0131l\u0131k, biyolojik sistemlerin \u00f6zellikleriyle zihnin, bilinci de i\u00e7eren \u00f6zellikleri aras\u0131nda s\u0131k\u0131 ba\u011flant\u0131lar oldu\u011funu vurgulayan bir d\u00fc\u015f\u00fcnce ailesine aittir.<\/strong> John Searle taraf\u0131ndan ortaya konan \u00f6zg\u00fcn form\u00fclasyonunda, bilin\u00e7li durumlar\u0131n, daha alt d\u00fczeydeki n\u00f6robiyolojik durumlar\u0131n indirgenemez \u00fcst d\u00fczey \u00f6zellikleri oldu\u011fu ileri s\u00fcr\u00fclm\u00fc\u015ft\u00fcr<\/strong> (Searle, 2017).<\/strong> Ben biyolojik do\u011falc\u0131l\u0131\u011f\u0131 bug\u00fcn yayg\u0131nla\u015fm\u0131\u015f olsa da biraz daha farkl\u0131 bir \u015fekilde kullan\u0131yorum. Bilincin yaln\u0131zca ya\u015fayan sistemlerin (ancak zorunlu olarak hepsinin de\u011fil) bir \u00f6zelli\u011fi oldu\u011funu ileri s\u00fcren bir sav olarak…<\/strong> Biyolojik do\u011falc\u0131l\u0131\u011f\u0131 t\u00fcm<\/em> canl\u0131 sistemlerin bilin\u00e7li oldu\u011funu ileri s\u00fcren \u201cbiyopsi\u015fizm\u201dden ay\u0131rmak gerekmektedir (Haeckel, 1892; Thompson, 2022).<\/p>\n\n\n\n

<\/div>\n\n\n\n

4.1 \u00d6ng\u00f6r\u00fcc\u00fc i\u015flemeye dayal\u0131 yakla\u015f\u0131m<\/em><\/strong><\/p>\n\n\n\n

<\/div>\n\n\n\n

Benim anlad\u0131\u011f\u0131m \u015fekliyle \u00f6ng\u00f6r\u00fcc\u00fc i\u015fleme [predictive processing<\/em>], alg\u0131, bili\u015f ve eylemle ilgili bir dizi kuram\u0131n ad\u0131d\u0131r (Clark, 2013; Hohwy, 2013; Hohwy & Seth, 2020; Rao & Ballard, 1999). Bu kuramlar\u0131n ortak noktas\u0131, alg\u0131n\u0131n olas\u0131l\u0131ksal bir \u00e7\u0131kar\u0131m bi\u00e7imi oldu\u011fu ve n\u00f6ral sistemlerde bunun \u00f6ng\u00f6r\u00fc hatas\u0131n\u0131 en aza indirme [prediction error minimisation] yoluyla yakla\u015f\u0131k olarak ger\u00e7ekle\u015fti\u011fi fikridir.<\/strong><\/p>\n\n\n\n

Standart anlat\u0131, alg\u0131sal i\u00e7eri\u011fin, d\u0131\u015far\u0131dan gelen duyu sinyallerinden \u201cokunarak\u201d [read out<\/em>] elde edilmedi\u011fini s\u00f6yler. Aksine i\u00e7erik, beynin bu sinyallerin nedenleri hakk\u0131nda olu\u015fturdu\u011fu \u201cen iyi tahmin\u201d taraf\u0131ndan verilmektedir. Bayes\u00e7i terimlerle ifade edilirse alg\u0131sal i\u00e7erikler, \u00f6nsel [prior<\/em>] inan\u00e7larla duyu sinyallerinin (olabilirliklerin [likelihood<\/em>]) a\u011f\u0131rl\u0131kl\u0131 birle\u015fiminden t\u00fcreyen yakla\u015f\u0131k olarak optimal ard\u0131l [posterior<\/em>] inan\u00e7lara kar\u015f\u0131l\u0131k gelir.[24]<\/a> Tam Bayes\u00e7i \u00e7\u0131kar\u0131m genellikle analitik olarak \u00e7\u00f6z\u00fclemez oldu\u011fundan, beyin, \u00f6ng\u00f6r\u00fc hatas\u0131n\u0131 en aza indirmeye y\u00f6nelik bir kestirim [approximation<\/em>] uygular. Buna g\u00f6re beyin, \u00fcretici modellerden olu\u015fan bir hiyerar\u015fi (Bayes\u00e7i terimlerle \u00f6nsel inan\u00e7 ile olabilirli\u011fin ortak da\u011f\u0131l\u0131m\u0131) uygular ki bu modeller, beklenen duyu sinyallerine ili\u015fkin alg\u0131sal hiyerar\u015filer boyunca a\u015fa\u011f\u0131 do\u011fru yay\u0131lan \u00f6ng\u00f6r\u00fcler \u00fcretir. \u201cYukar\u0131dan a\u015fa\u011f\u0131\u201d (i\u00e7ten d\u0131\u015fa) \u00f6ng\u00f6r\u00fcler, \u201ca\u015fa\u011f\u0131dan yukar\u0131\u201d (d\u0131\u015ftan i\u00e7e) duyu sinyalleriyle kar\u015f\u0131la\u015ft\u0131r\u0131l\u0131r ve \u00f6ng\u00f6r\u00fc hatalar\u0131 ortaya \u00e7\u0131kar. Bu hatalar hiyerar\u015fi boyunca yeniden yukar\u0131 iletilir. \u00d6ng\u00f6r\u00fcleri g\u00fcncelleyip hatalar\u0131 en aza indiren bu s\u00fcre\u00e7, yakla\u015f\u0131k olarak optimal bir ard\u0131l, yani \u201cen iyi tahmin\u201d \u00fczerinde karar k\u0131lar. \u00d6ng\u00f6r\u00fc hatalar\u0131, beklenen duyu verisini ortaya \u00e7\u0131karacak eylemler ger\u00e7ekle\u015ftirerek de azalt\u0131labilir ki buna da \u201caktif \u00e7\u0131kar\u0131m\u201d [active inference<\/em>] denir (Friston ve di\u011ferleri, 2017).<\/p>\n\n\n\n

Bilin\u00e7 a\u00e7\u0131s\u0131ndan temel iddia, bilin\u00e7li i\u00e7eriklerin, beynin \u201cen iyi tahminleri\u201d taraf\u0131ndan verildi\u011fidir; bir di\u011fer deyi\u015fle, yukar\u0131dan a\u015fa\u011f\u0131 alg\u0131sal \u00f6ng\u00f6r\u00fclerin ortak i\u00e7eri\u011fi bilinci belirler (Hohwy & Seth, 2020; Whyte ve di\u011ferleri, 2024). Bu sezgi, bilin\u00e7li alg\u0131n\u0131n \u201cdenetimli hal\u00fcsinasyon\u201d oldu\u011fu d\u00fc\u015f\u00fcncesinde ifadesini bulur ve bilin\u00e7li i\u00e7eriklerin pasif olarak kaydedilmek yerine etkin bi\u00e7imde \u00fcretildi\u011fini vurgular (Seth, 2021).[25]<\/a><\/p>\n\n\n\n

Bu standart anlat\u0131, d\u00fcnyan\u0131n yan\u0131 s\u0131ra bedenin bilin\u00e7li alg\u0131lar\u0131n\u0131 da a\u00e7\u0131klayacak \u015fekilde geni\u015fletilebilir. Bedenle ilgili alg\u0131lar, \u201cbenlik\u201d deneyiminin vazge\u00e7ilmez bir par\u00e7as\u0131n\u0131 olu\u015fturur. Benlik a\u00e7\u0131s\u0131ndan merkezi g\u00f6r\u00fcnen duygu ve ruh h\u00e2li deneyimleri, bedenin fizyolojik durumunu bildiren duyu girdilerinin nedenleri hakk\u0131nda olas\u0131l\u0131ksal en iyi tahminler anlam\u0131na gelen interoseptif (i\u00e7duyumsal) \u00e7\u0131kar\u0131mlara ba\u011fl\u0131 olabilir<\/strong> (Barrett & Simmons, 2015; Seth, 2013).<\/strong> \u00d6nemli olan nokta, interoseptif \u00e7\u0131kar\u0131m\u0131n i\u015flevsel rol\u00fcn\u00fcn, neyin orada oldu\u011funu ke\u015ffetmekten \u00e7ok fizyolojik d\u00fczenleme ile ilgili olmas\u0131d\u0131r.<\/strong> Bu ayr\u0131m, ara\u00e7sal [instrumental<\/em>] \u00e7\u0131kar\u0131m ile epistemik \u00e7\u0131kar\u0131m aras\u0131ndaki fark a\u00e7\u0131s\u0131ndan ortaya konulabilir. Epistemik \u00e7\u0131kar\u0131m, standart \u00f6ng\u00f6r\u00fcc\u00fc i\u015flemeye uygun bi\u00e7imde, duyu sinyallerinin en olas\u0131 nedenlerini \u00e7\u0131karsamaya y\u00f6neliktir. Ara\u00e7sal \u00e7\u0131kar\u0131m ise ke\u015ffetmeye de\u011fil denetime \u00f6ncelik verir. Ara\u00e7sal \u00e7\u0131kar\u0131mda \u00f6nsel inan\u00e7lar hedef ya da ayar noktas\u0131 i\u015flevi g\u00f6rebilir. Aktif \u00e7\u0131kar\u0131m yoluyla (interoseptif) \u00f6ng\u00f6r\u00fc hatas\u0131n\u0131 en aza indirmek, bu ayar noktalar\u0131yla ili\u015fkili homeostatik ya da allostatik d\u00fczenlemeleri uygulamaya koyabilir (Seth & Friston, 2016).[26]<\/a><\/p>\n\n\n\n

\u0130nteroseptif \u00e7\u0131kar\u0131ma ili\u015fkin bu denetim odakl\u0131 bak\u0131\u015f, denetim i\u00e7in \u00f6ng\u00f6r\u00fcn\u00fcn \u00f6nemini vurgulayan ve k\u00f6kl\u00fc bir ge\u00e7mi\u015fe sahip sibernetik d\u00fc\u015f\u00fcnceleri miras alm\u0131\u015ft\u0131r (Conant & Ashby, 1970; Seth, 2015). Bu durum, bilin\u00e7li alg\u0131n\u0131n genel bir mekanizmas\u0131 olarak \u00f6ng\u00f6r\u00fcc\u00fc i\u015fleme y\u00f6n\u00fcndeki evrimsel itkinin, allostaz i\u00e7in temel bir biyolojik zorunluluk olabilece\u011fini d\u00fc\u015f\u00fcnd\u00fcr\u00fcr. K\u0131sacas\u0131 hayatta kalma zorunlulu\u011fu…<\/strong> Ayr\u0131ca baz\u0131 fenomenolojik ayr\u0131mlar\u0131 da ayd\u0131nlat\u0131r. G\u00f6rsel deneyimler, ilk kestirim olarak, uzamsall\u0131k ve nesneye y\u00f6nelim gibi bir dizi fenomenolojik \u00f6zelli\u011fi payla\u015f\u0131r. Bu \u00f6zellikler, genel i\u015flevi \u201cneyin orada oldu\u011funu\u201d bulmak olan (epistemik) \u00f6ng\u00f6r\u00fclerle \u00f6rt\u00fc\u015f\u00fcr. Duygusal deneyimler daha farkl\u0131 bir karakter ta\u015f\u0131r, en ba\u015fta de\u011ferlilik [valence<\/em>], yani iyi ya da k\u00f6t\u00fc hissetmenin \u00e7e\u015fitleri taraf\u0131ndan \u015fekillenirler. Bu \u00f6zellikler, genel i\u015flevi allostatik d\u00fczenleme olan denetim odakl\u0131 \u00f6ng\u00f6r\u00fclerle uyumludur.<\/p>\n\n\n\n

Bu g\u00f6zlemler bir araya geldi\u011finde, insan\u0131n bilin\u00e7li deneyimlerinin ancak ya\u015fayan ve kendini s\u00fcrd\u00fcren organizmalar\u0131n do\u011fas\u0131 \u0131\u015f\u0131\u011f\u0131nda anla\u015f\u0131labilece\u011fini s\u00f6yleyen bir t\u00fcr epistemik<\/em> biyolojik do\u011falc\u0131l\u0131\u011fa i\u015faret eder.<\/strong>[27]<\/a> Bu iddia, d\u00fcnyaya (ve bedene) genel olarak duyarl\u0131l\u0131ktan ziyade, bilin\u00e7li alg\u0131 i\u00e7in ge\u00e7erlidir. \u00c7\u00fcnk\u00fc a\u00e7\u0131klama, bilin\u00e7li deneyimin fenomenolojik \u00f6zellikleriyle kurulan ba\u011flara dayan\u0131r (Seth, 2016). Hatta bilin\u00e7li deneyimin asgar\u00ee bir \u201czemin durumu\u201dnun bulundu\u011fu ileri s\u00fcr\u00fclebilir. B\u00fct\u00fcn\u00fcyle i\u00e7eri\u011fi olmayan bir h\u00e2l yerine, yaln\u0131zca \u201cya\u015f\u0131yor olma\u201dn\u0131n r\u00fc\u015feym h\u00e2lindeki, \u015fekilsiz ve bi\u00e7imsiz bir hissi s\u00f6z konusu olabilir.<\/strong>[28]<\/a> Bu varsay\u0131msal zemin durum, ister benli\u011fe ister d\u00fcnyaya ili\u015fkin olsun, di\u011fer t\u00fcm bilin\u00e7li i\u00e7erik bi\u00e7imleri i\u00e7in bir iskele kurabilir [scaffold<\/em>].<\/strong> Serbest\u00e7e ifade edersek, d\u00fcnyay\u0131 ve benli\u011fimizi, ya\u015fayan bedenlerimizle birlikte<\/em>, onlar\u0131n arac\u0131l\u0131\u011f\u0131yla<\/em> ve onlar sayesinde<\/em> deneyimleriz<\/strong> (Seth, 2021; ayr\u0131ca bkz. Seth, 2019; Seth & Tsakiris, 2018).<\/strong><\/p>\n\n\n\n

“D\u00fcnyay\u0131 ve benli\u011fimizi, ya\u015fayan bedenlerimizle birlikte<\/em>, onlar\u0131n arac\u0131l\u0131\u011f\u0131yla<\/em> ve onlar sayesinde<\/em> deneyimleriz.”<\/p><\/blockquote><\/figure>\n\n\n\n

Bir ad\u0131m daha ileri gidersek hayatta kalma zorunlulu\u011funun biyolojimizde belirli bir \u201cuygulama d\u00fczeyinde\u201d son bulmad\u0131\u011f\u0131n\u0131 g\u00f6r\u00fcr\u00fcz.<\/strong> N\u00f6ronlar d\u00e2hil tek h\u00fccreler bile s\u00fcrekli kendi fizyolojik b\u00fct\u00fcnl\u00fcklerini s\u00fcrd\u00fcrmekle me\u015fguld\u00fcr. Bu homeostatik s\u00fcre\u00e7, daha \u00fcst d\u00fczey i\u015flevsel \u00f6zelliklerle do\u011frudan ba\u011flant\u0131land\u0131r\u0131labilir. Daha \u00f6nce verdi\u011fim \u00f6rne\u011fi hat\u0131rlayal\u0131m. G\u00f6r\u00fcn\u00fc\u015fte kendili\u011finden olan aksiyon potansiyeli de\u015farjlar\u0131n\u0131n zararl\u0131 kimyasallar\u0131n birikimini \u00f6nleyerek n\u00f6ronlar\u0131 koruyabildi\u011fi, dolay\u0131s\u0131yla n\u00f6ronal ate\u015fleme \u00f6r\u00fcnt\u00fclerinin k\u0131smen bu s\u00fcrece ba\u011fl\u0131 olabilece\u011fi g\u00f6sterilmi\u015ftir (Chintaluri & Vogels, 2023). Substrat ile ondan kaynaklanan s\u00fcre\u00e7ler ve dinamikler aras\u0131nda net bir s\u0131n\u0131r \u00e7izilemiyorsa, substrat\u0131n \u00f6nemsiz oldu\u011fu<\/em> y\u00f6n\u00fcndeki sezgiyi savunmak, ki bu sezgi i\u015flemlemeselci i\u015flevselcili\u011fin merkezindedir, daha g\u00fc\u00e7 h\u00e2le gelir.<\/strong><\/p>\n\n\n\n

Bir ba\u015fka yakla\u015f\u0131m, ya\u015fayan sistemlerin \u201cellerini ta\u015f\u0131n alt\u0131na koydu\u011funu\u201d [skin in the game<\/em>] s\u00f6yler. \u00d6rg\u00fctlenmenin birden \u00e7ok d\u00fczeyi i\u00e7inde ve d\u00fczeyler aras\u0131nda kendi kendini idame s\u00fcre\u00e7lerini kesintisiz bir \u015fekilde y\u00fcr\u00fct\u00fcrler (Aru ve di\u011ferleri, 2023). Bu g\u00f6r\u00fc\u015fe g\u00f6re \u00f6ng\u00f6r\u00fcc\u00fc i\u015flemenin (\u00fcst d\u00fczey) mekanizmalar\u0131, ayr\u0131lmaz bi\u00e7imde (alt d\u00fczey) biyolojik ba\u011flamlar\u0131na g\u00f6m\u00fcl\u00fcd\u00fcr. Bu g\u00f6m\u00fclme, \u00fcst d\u00fczey \u00f6zelliklerin ger\u00e7ekle\u015fmesi i\u00e7in gereklidir ve \u00f6nceki arg\u00fcmanlar\u0131n da g\u00f6sterdi\u011fi \u00fczere bilin\u00e7le ba\u011flant\u0131lar\u0131n\u0131 a\u00e7\u0131klamaya yard\u0131mc\u0131 olur.<\/strong><\/p>\n\n\n\n

Buradan Yunanca \u201c\u00f6z-\u00fcretim\u201d veya \u201c\u00f6z-yarat\u0131m\u201d anlam\u0131ndan t\u00fcreyen otopoiesis<\/em><\/strong> kavram\u0131na geliyoruz. Otopoietik sistemler bir s\u00fcre\u00e7ler a\u011f\u0131 arac\u0131l\u0131\u011f\u0131yla kendi maddi bile\u015fenlerini durmaks\u0131z\u0131n yeniden \u00fcretmeleri ve de bu s\u00fcre\u00e7ler a\u011f\u0131n\u0131n, sistem ile \u00e7evresi aras\u0131ndaki s\u0131n\u0131r\u0131 etkin bi\u00e7imde s\u00fcrd\u00fcrmesi hasebiyle \u00f6zeldirler <\/strong>(Maturana & Varela, 1980).<\/strong> Biyolojik h\u00fccreler, otopoietik sistemlerin ba\u015fat \u00f6rnekleridir. Yaln\u0131zca i\u015flevleri yerine getirmezler, ayn\u0131 zamanda kendi maddi temellerini (\u00f6z-\u00fcretim) ve b\u00fct\u00fcnl\u00fcklerini (\u00f6z-kimlik) s\u00fcrekli yeniden \u00fcretirler.<\/strong> Bilgisayarlar ve fabrikalar gibi insan tasar\u0131m\u0131 sistemler genellikle otopoietik de\u011fildir. Bir fabrika hammadde al\u0131r ve s\u00f6zgelimi otomobil gibi bir \u00e7\u0131kt\u0131 \u00fcretir. Fabrika kendini \u00fcretmez. Bir bilgisayar girdi al\u0131r ve \u00e7\u0131kt\u0131 \u00fcretir. O da kendini \u00fcretmez. Bu t\u00fcr sistemler allopoietik<\/em><\/strong>, yani \u201cba\u015fkas\u0131nca-\u00fcretimli\u201ddir [others producing<\/em>]. Otopoietik sistemlerin ne olduklar\u0131n\u0131 ne yapt\u0131klar\u0131ndan ay\u0131rman\u0131n ger\u00e7ekte hi\u00e7bir yolu yoktur.<\/strong>[29]<\/a><\/p>\n\n\n\n

Bu noktada ortaya \u00e7\u0131kan d\u00fc\u015f\u00fcnce, ya\u015fam ile bilin\u00e7 aras\u0131nda daha g\u00fc\u00e7l\u00fc bir s\u00fcreklilik bulundu\u011funa, t\u00fcm bilin\u00e7li deneyimlerin temelindeki \u00f6ng\u00f6r\u00fcc\u00fc s\u00fcre\u00e7lerin, canl\u0131 maddenin otopoietik ve uyar\u0131labilir do\u011fas\u0131na ayr\u0131lmaz bi\u00e7imde ba\u011fl\u0131 oldu\u011funa i\u015faret eder.<\/strong>[30]<\/a> Bu fikrin motivasyonu, h\u00fccresel otopoiesisin bizzat bir t\u00fcr \u00f6ng\u00f6r\u00fc hatas\u0131n\u0131 en aza indirmeye ba\u011fl\u0131 olarak anla\u015f\u0131labilece\u011finin ya da k\u0131smen bu t\u00fcrde bir en aza indirmeden olu\u015ftu\u011funun fark edilmesine dayan\u0131r. Bu da bizi serbest enerji ilkesi<\/em>yle bulu\u015fturur<\/strong>.<\/p>\n\n\n\n

“Otopoietik sistemlerin ne olduklar\u0131n\u0131 ne yapt\u0131klar\u0131ndan ay\u0131rman\u0131n ger\u00e7ekte hi\u00e7bir yolu yoktur.”<\/p><\/blockquote><\/figure>\n\n\n\n

<\/div>\n\n\n\n

4.2 Serbest enerji ilkesine dayal\u0131 yakla\u015f\u0131m<\/em><\/p>\n\n\n\n

<\/div>\n\n\n\n

\u00d6nce \u00f6ng\u00f6r\u00fcc\u00fc i\u015fleme anlat\u0131s\u0131n\u0131 ele ald\u0131\u011f\u0131m\u0131za g\u00f6re, d\u00f6n\u00fc\u015f yolculu\u011fumuz daha h\u0131zl\u0131 ilerleyebilir. Bu yolculuk, varl\u0131\u011f\u0131n\u0131 s\u00fcrd\u00fcrmek i\u00e7in d\u00fczensizli\u011fe etkin bi\u00e7imde kar\u015f\u0131 koyan sistemlerin do\u011fas\u0131na ili\u015fkin temel k\u0131s\u0131tlar\u0131n belirlenmesiyle ba\u015flar ve buradan bilince ili\u015fkin daha geni\u015f \u00f6ng\u00f6r\u00fcc\u00fc i\u015fleme g\u00f6r\u00fc\u015f\u00fcne yeniden ba\u011flan\u0131r.<\/strong>[31]<\/a><\/p>\n\n\n\n

Ba\u015flang\u0131\u00e7 noktas\u0131, h\u00fccrelerden organizman\u0131n b\u00fct\u00fcn\u00fcne, ya\u015fayan sistemlerin otopoietik do\u011fas\u0131d\u0131r.<\/strong> Kayalardan ya da bilgisayar gibi di\u011fer \u015feylerden farkl\u0131 olarak bu sistemler kendi maddi temellerini etkin bi\u00e7imde (yeniden) \u00fcretir ve zaman boyunca s\u0131n\u0131rlar\u0131n\u0131 korur. Serbest enerji ilkesi a\u00e7\u0131s\u0131ndan bu, i\u00e7 durumlar\u0131n\u0131n da\u011f\u0131lmas\u0131na [dispersion<\/em>] etkin bi\u00e7imde kar\u015f\u0131 koyduklar\u0131 anlam\u0131na gelir. Kar\u015f\u0131 koyulmad\u0131\u011f\u0131 takdirde bu da\u011f\u0131lma, termodinami\u011fin ikinci yasas\u0131 uyar\u0131nca zorunlu olarak ger\u00e7ekle\u015fir. Bu nedenle \u00e7evreleriyle termodinamik dengeden uzak, d\u00fc\u015f\u00fck entropili h\u00e2llerde bulunmalar\u0131 ve ka\u00e7\u0131n\u0131lmaz son olan d\u00fczensizli\u011fe s\u00fcr\u00fcklenmeyi savu\u015fturmalar\u0131 gerekir.<\/strong> Hayatta kalmak, entropiyi s\u00fcrekli olarak en aza indirmek demektir, \u00e7\u00fcnk\u00fc bir bulamaca [mush<\/em>] d\u00f6n\u00fc\u015fmenin yollar\u0131, ya\u015faman\u0131n yollar\u0131ndan \u00e7ok daha fazlad\u0131r.<\/strong>

Bazen hayatta kalman\u0131n termodinami\u011fin ikinci yasas\u0131na kar\u015f\u0131 koymay\u0131 gerektirdi\u011fi s\u00f6ylenir. Bunu, ya\u015fayan sistemlerin ikinci yasadan yararland\u0131\u011f\u0131<\/em> \u015feklinde ifade etmek daha yerindedir. \u00c7\u00fcnk\u00fc canl\u0131lar\u0131n dengeden uzak (yar\u0131) kararl\u0131 durumlarda kalabilmesini sa\u011flayan, metabolizma yoluyla d\u00fc\u015f\u00fck entropili yak\u0131t\u0131n\/besinin y\u00fcksek entropili \u00fcr\u00fcnlere d\u00f6n\u00fc\u015ft\u00fcr\u00fclmesidir.<\/strong>[32]<\/a><\/p>\n\n\n\n

Ortaya \u00e7\u0131kan g\u00fc\u00e7l\u00fck \u015fu ki, sistem, entropiyi do\u011frudan \u00f6l\u00e7emez. Dolay\u0131s\u0131yla entropi (do\u011frudan) en aza indirilemez. Bu noktada serbest enerji devreye girer. Serbest enerji, ya da teknik ifadeyle varyasyonel serbest enerji<\/em>, entropi i\u00e7in \u00f6l\u00e7\u00fclebilir bir \u00fcst s\u0131n\u0131r sa\u011flar.<\/strong>[33]<\/a> Bu nedenle (varyasyonel) serbest enerjinin en aza indirilmesi, entropinin de azalmas\u0131na yol a\u00e7ar (Buckley ve di\u011ferleri, 2017; Parr ve di\u011ferleri, 2022). \u201c\u00d6l\u00e7\u00fclebilir\u201dden kas\u0131t, serbest enerjinin duyu sinyallerinin bir fonksiyonu olmas\u0131 ve ilkece sistem taraf\u0131ndan eri\u015filebilmesidir. Uygulamada, baz\u0131 yal\u0131n matematiksel varsay\u0131mlar alt\u0131nda serbest enerji, duyuya ili\u015fkin bir \u00f6ng\u00f6r\u00fc hatas\u0131 olarak yorumlanabilir.[34]<\/a> Bu durumda duyuya ili\u015fkin \u00f6ng\u00f6r\u00fc hatalar\u0131n\u0131 etkin bi\u00e7imde en aza indirmek, kuramsal olarak daha temel bir nicelik olan serbest enerjiyi de en aza indirmeyi gerektirecektir.<\/p>\n\n\n\n

Tersine \u00e7evirirsek, serbest enerjinin en aza indirilmesi, ayn\u0131 varsay\u0131mlar alt\u0131nda, bir sistemin \u00fcretici modellerinin kodlad\u0131\u011f\u0131 yakla\u015f\u0131k (Bayes\u00e7i) ard\u0131l inan\u00e7lar\u0131n optimal ard\u0131l inan\u00e7lara yakla\u015fmas\u0131na yol a\u00e7ar, ayn\u0131 zamanda bu modellere ili\u015fkin duyusal kan\u0131t\u0131n da azamiye \u00e7\u0131kar\u0131lmas\u0131na hizmet eder.[35]<\/a> Bu \u00e7er\u00e7eve, serbest enerji ilkesi ile \u00f6ng\u00f6r\u00fcc\u00fc i\u015fleme aras\u0131ndaki s\u00fcreklili\u011fin alt\u0131n\u0131 \u00e7izer.<\/p>\n\n\n\n

Buradaki genel sezgi,ya\u015fayan sistemlerin aktif \u00e7\u0131kar\u0131m yoluyla duyusal \u00f6ng\u00f6r\u00fc hatas\u0131n\u0131 en aza indirerek, kendilerinin do\u011fal olarak i\u00e7inde bulunmay\u0131 bekledikleri (ya da \u00f6ng\u00f6rd\u00fckleri) durumlarda kalma e\u011filimi g\u00f6stermeleri ve b\u00f6ylece varl\u0131klar\u0131n\u0131 s\u00fcrd\u00fcrmeleridir.<\/strong> Ba\u015fka bir deyi\u015fle, olduklar\u0131 t\u00fcrden \u015feyler olarak kal\u0131rlar.<\/strong> Jakob Hohwy bu sezgiyi, organizmalar\u0131n kendi varolu\u015flar\u0131 lehine kan\u0131t\u0131 azami k\u0131ld\u0131klar\u0131n\u0131 s\u00f6yleyerek derli toplu bi\u00e7imde dile getirir, organizmalar \u201ckendi kendilerinin kan\u0131t\u0131\u201d [self-evidence<\/em>] olmaktad\u0131r<\/strong> (Hohwy, 2014).<\/strong><\/p>\n\n\n\n

Serbest enerji ilkesinin bu y\u00f6nleri, aktif \u00e7\u0131kar\u0131m\u0131n t\u00fcm yap\u0131s\u0131n\u0131n ya\u015fayan sistemlerin otopoietik do\u011fas\u0131na kadar uzand\u0131\u011f\u0131n\u0131 d\u00fc\u015f\u00fcnd\u00fcr\u00fcr.<\/strong> Bu s\u00fcreklilik, metabolizmada termodinamik serbest enerjinin (en aza indirilmesi) ile \u00f6ng\u00f6r\u00fcc\u00fc i\u015flemeye dayal\u0131 yakla\u015f\u0131mlarda varyasyonel serbest enerjinin (en aza indirilmesi) aras\u0131nda bir e\u015fde\u011ferlik bulundu\u011funun kaydedilmesiyle daha da g\u00fc\u00e7lenir. Metabolik ba\u011flamdaki en aza indirme, otopoietik b\u00fct\u00fcnl\u00fc\u011f\u00fc ve dinamik bir kendilik\/ba\u015fkas\u0131 s\u0131n\u0131r\u0131n\u0131n s\u00fcrd\u00fcr\u00fclmesini temellendirir. Serbest enerji ilkesi yaz\u0131n\u0131nda bu s\u0131n\u0131ra \u201cMarkov \u00f6rt\u00fcs\u00fc\u201d denir (Friston, 2013; ayr\u0131ca bkz. Di Paolo ve di\u011ferleri, 2022; Nave, 2025). \u00d6ng\u00f6r\u00fcc\u00fc i\u015flemeye dayal\u0131 en aza indirme ise bilin\u00e7li deneyimin \u00f6ng\u00f6r\u00fcc\u00fc i\u015fleme ve denetimli hal\u00fcsinasyon \u00fczerinden anla\u015f\u0131lmas\u0131n\u0131 g\u00fc\u00e7lendirir.<\/p>\n\n\n\n

Jaynes\u2019in (1957) \u00e7al\u0131\u015fmalar\u0131na dayanan fizikte geni\u015f kabul g\u00f6rm\u00fc\u015f iki sonu\u00e7, bu e\u015fde\u011ferli\u011fi daha da derinle\u015ftirir. Jarzynski e\u015fitli\u011fi, durumlar aras\u0131 ge\u00e7i\u015f s\u0131ras\u0131nda yap\u0131lan ortalama termodinamik i\u015fin, bu durumlar aras\u0131ndaki serbest enerji fark\u0131 taraf\u0131ndan s\u0131n\u0131rland\u0131\u011f\u0131n\u0131 g\u00f6sterir (Jarzynski, 1997). Bu sonu\u00e7, \u00f6zellikle ya\u015fayan sistemler gibi dengeden uzak s\u00fcre\u00e7lere uygulan\u0131r ve metabolizma gibi s\u00fcre\u00e7lere serbest enerji temelli bir form\u00fclasyon kazand\u0131r\u0131r. Jarzynski e\u015fitli\u011fi, bir bit bilginin silinmesi i\u00e7in gereken (termodinamik) enerjinin alt s\u0131n\u0131r\u0131n\u0131 veren Landauer ilkesinin bir genellemesi olarak g\u00f6r\u00fclebilir (Landauer, 1961). Bu s\u0131n\u0131r, \u00f6ng\u00f6r\u00fcc\u00fc i\u015flemede \u00e7\u0131kar\u0131m\u0131n asgari bir termodinamik maliyeti oldu\u011funu ima eder. Daha yak\u0131n zamanda Sengupta ve \u00e7al\u0131\u015fma arkada\u015flar\u0131, aktif \u00e7\u0131kar\u0131m ba\u011flam\u0131nda varyasyonel serbest enerjinin en aza indirilmesinin hem metabolik hem de istatistiksel (\u00e7\u0131kar\u0131msal) verimlili\u011fe yol a\u00e7t\u0131\u011f\u0131n\u0131 g\u00f6sterdiler (Sengupta ve di\u011ferleri, 2013; ayr\u0131ca bkz. Fields ve di\u011ferleri, 2024). Buradaki cazip fikir, \u00f6ng\u00f6r\u00fcler do\u011fru oldu\u011funda, termodinamik bak\u0131mdan maliyetli durum de\u011fi\u015fikliklerine gerek kalmad\u0131\u011f\u0131d\u0131r.<\/p>\n\n\n\n

Hepsi bir araya getirildi\u011finde, metabolizma, otopoiesis ve alg\u0131y\u0131 y\u00fcr\u00fcten serbest enerjinin (en aza indirilmesinin) yaln\u0131zca metaforik bir e\u015fde\u011ferlik olmay\u0131p fiziksel bir anlamda da ayn\u0131 s\u00fcre\u00e7 oldu\u011fu sonucuna i\u015faret eder. Bununla birlikte bu e\u015fde\u011ferli\u011fin tam olarak ne anlama geldi\u011fi h\u00e2l\u00e2 tart\u0131\u015fmal\u0131d\u0131r <\/strong>(Collell & Fauquet, 2015).<\/strong> Serbest enerji ilkesinin kendisi de, en az\u0131ndan daha geni\u015f iddialar\u0131 ve otopoiesisle ili\u015fkisi bak\u0131m\u0131ndan, yo\u011fun bi\u00e7imde tart\u0131\u015f\u0131lmaktad\u0131r (Aguilera ve di\u011ferleri, 2022; Biehl ve di\u011ferleri, 2021; Bruineberg ve di\u011ferleri, 2021; Di Paolo ve di\u011ferleri, 2022; Nave, 2025).<\/p>\n\n\n\n

Bu noktada a\u015fa\u011f\u0131daki gibi bir kayg\u0131 dile getirilebilir. \u00d6ng\u00f6r\u00fcc\u00fc i\u015fleme, \u00f6ng\u00f6r\u00fcn\u00fcn ve \u00f6ng\u00f6r\u00fc hatas\u0131n\u0131n Bayes\u00e7i dans\u0131yla beraber, paradigmatik bir i\u015flemleme s\u00fcreci \u00f6rne\u011fi gibi g\u00f6r\u00fcnmektedir.<\/strong> Elbette \u00f6ng\u00f6r\u00fcc\u00fc i\u015fleme (ve Bayes\u00e7i \u00e7\u0131kar\u0131m) bir algoritma olarak kullan\u0131labilir ve s\u0131k s\u0131k da b\u00f6yle kullan\u0131l\u0131r. Kuram, (bilin\u00e7li) alg\u0131sal i\u00e7eriklerin, olas\u0131l\u0131ksal \u00e7\u0131kar\u0131m s\u00fcreci \u015feklinde faydal\u0131 bi\u00e7imde modellenebilece\u011fini<\/em> \u00f6ne s\u00fcr\u00fcyor gibi de yorumlanabilir (Andrews, 2021). Ancak serbest enerji ilkesiyle kurulan s\u00fcreklilik \u00f6nemli bir ayr\u0131ma i\u015faret eder. Biyolojik sistemlerde \u00f6ng\u00f6r\u00fcc\u00fc i\u015fleme, s\u00fcrekli bir niceli\u011fin (serbest enerjinin) en aza indirilmesine dayanan, dinamik ve muhtemelen substrata ba\u011f\u0131ml\u0131 bir s\u00fcre\u00e7tir.<\/strong> Bu perspektif, alg\u0131sal i\u00e7eriklerin yaln\u0131zca bu yolla modellenmekle kalmay\u0131p, bizzat alg\u0131sal \u00e7\u0131kar\u0131m\u0131n baz\u0131 y\u00f6nleri taraf\u0131ndan<\/em> kuruldu\u011funu<\/em> [constituted by<\/em>] ileri s\u00fcren daha g\u00fc\u00e7l\u00fc iddiaya k\u0131smen zemin sa\u011flar. Bu anlamda \u00f6ng\u00f6r\u00fcc\u00fc i\u015fleme, b\u00f6l\u00fcm 3.6.\u2019da \u00f6ng\u00f6r\u00fcld\u00fc\u011f\u00fc gibi dinamik, i\u015flemlemeselci olmayan bir kuram h\u00e2line gelir. Bu s\u00fcre\u00e7 i\u015flemlemesel olarak soyutlanabilir, fakat bu algoritmalar, biyolojik beyin ve bedenlerde a\u00e7\u0131\u011fa \u00e7\u0131kan bedenlenmi\u015f ve g\u00f6m\u00fcl\u00fc s\u00fcre\u00e7lerin solgun g\u00f6lgeleridir.<\/strong><\/p>\n\n\n\n

<\/div>\n\n\n\n

4.3 Zarlar, mitokondriler ve zenobotlar<\/em><\/p>\n\n\n\n

<\/div>\n\n\n\n

Yeni ortaya \u00e7\u0131kan di\u011fer \u00e7al\u0131\u015fma hatlar\u0131, genel olarak biyolojik i\u015flevlerin ve \u00f6zellikle bilincin substrata ba\u011f\u0131ml\u0131l\u0131\u011f\u0131 hakk\u0131nda ilgi \u00e7ekici ipu\u00e7lar\u0131 sunar.<\/strong> \u0130ki tanesinden s\u00f6z edece\u011fim. Birincisi \u201chissetme\u201d i\u00e7in fiziksel bir temelin bakteriler gibi basit canl\u0131 sistemlerde bile bulunabilece\u011fini savunan Nick Lane\u2019den gelmektedir.<\/strong> Lane, metabolizma yoluyla dinamik bi\u00e7imde s\u00fcrd\u00fcr\u00fclen bakteri zar\u0131 boyunca olu\u015fan elektriksel potansiyellerin, organizman\u0131n fizyolojik durumunu \u00e7evresiyle ili\u015fkisi i\u00e7inde b\u00fct\u00fcnle\u015fik, de\u011fer ta\u015f\u0131yan [valenced<\/em>] ve eylem y\u00f6nelimli bir \u00e7\u0131kt\u0131 olarak sundu\u011funu \u00f6ne s\u00fcrer <\/strong>(Lane, 2022).<\/strong>[36]<\/a> G\u00f6r\u00fcld\u00fc\u011f\u00fc gibi, serbest enerji ilkesiyle ba\u011flant\u0131lar mevcuttur, nitekim bakteri zar\u0131 bir Markov \u00f6rt\u00fcs\u00fc olarak d\u00fc\u015f\u00fcn\u00fclebilir. Ayr\u0131ca otopoiesisle de ba\u011flant\u0131lar s\u00f6z konusudur, zira metabolizma hem zar potansiyelini \u00fcretir hem de ondan g\u00fc\u00e7 al\u0131r.<\/p>\n\n\n\n

Lane, daha karma\u015f\u0131k canl\u0131larda bilin\u00e7li durumlar\u0131n olu\u015fumunda mitokondrilerin anahtar bir rol oynad\u0131\u011f\u0131n\u0131 \u00f6ne s\u00fcrer.<\/strong> Evrimsel olarak mitokondriler bakterilerden t\u00fcremi\u015ftir ve mitokondriyal plazma zar\u0131 bakteriyel zara benzer bir rol \u00fcstlenir. Metabolizman\u0131n ortaya \u00e7\u0131kard\u0131\u011f\u0131 ve s\u00fcrd\u00fcr\u00fclmesi i\u00e7in zorunlu olan \u015fa\u015f\u0131rt\u0131c\u0131 derecede g\u00fc\u00e7l\u00fc elektromanyetik alanlar\u0131n sinaptik etkinli\u011fi, n\u00f6ronal ate\u015flemeyi ve belki de EEG ile saptanabilen makroskopik elektromanyetik alanlar\u0131 bile etkileyebilece\u011fi savunulur. Ayr\u0131ca anesteziklerin bu alanlar\u0131n \u00fcretilmesinden sorumlu mitokondri i\u00e7indeki belirli s\u00fcre\u00e7leri bozarak etkili oldu\u011fu \u00f6ne s\u00fcr\u00fcl\u00fcr. Bu g\u00f6r\u00fc\u015fe g\u00f6re mitokondriyal (ve bakteriyel) i\u015flevin ilgili y\u00f6nleri bu organellerde ger\u00e7ekle\u015fen h\u00fccresel solunumun biyokimyas\u0131ndan soyutlanamaz.<\/strong><\/p>\n\n\n\n

\u0130kinci bak\u0131\u015f a\u00e7\u0131s\u0131 bizi Michael Levin\u2019in biyolojik maddenin \u201cfail\u201d [agential<\/em>] do\u011fas\u0131na dair yenilik\u00e7i g\u00f6r\u00fc\u015flerine g\u00f6t\u00fcr\u00fcr<\/strong> (Levin, 2023, 2025). Birlikte \u00e7al\u0131\u015ft\u0131\u011f\u0131 ekip, bu fikirleri \u201czenobot\u201d ve \u201cantrobot\u201d gibi tuhaf yarat\u0131mlar bi\u00e7iminde deneysel olarak ara\u015ft\u0131rm\u0131\u015ft\u0131r (Ebrahimkhani & Levin, 2021; Gumuskaya ve di\u011ferleri, 2024). Bunlar amfibi h\u00fccrelerin (zenobot) ya da insan h\u00fccrelerinin (antrobot) bir araya gelmesiyle olu\u015fan yap\u0131lard\u0131r ve h\u00fccrelerin al\u0131\u015f\u0131ld\u0131k ba\u011flamlar\u0131ndan ya da evrimsel tarihlerinden beklenebilece\u011finden \u00e7ok farkl\u0131 davran\u0131\u015fsal yeterliklere sahip yar\u0131-fail sistemlere kendiliklerinden \u00f6rg\u00fctlenirler (\u00f6rne\u011fin normalde hareketsiz olan insan soluk borusu h\u00fccrelerinden yap\u0131lm\u0131\u015f hareketli antrobotlar gibi).<\/strong> Burada hen\u00fcz bilin\u00e7le do\u011frudan bir ba\u011flant\u0131 bulunmamakla birlikte Levin\u2019in yarat\u0131mlar\u0131 biyolojik maddede i\u00e7kin olan i\u015flevsel ve dinamik potansiyelin derin da\u011farc\u0131\u011f\u0131n\u0131 ortaya \u00e7\u0131kar\u0131r. Bu bulgular genetik bilgi depolamay\u0131, kodlamay\u0131 ve kod \u00e7\u00f6zmeyi \u00f6nceliklendiren geleneksel g\u00f6r\u00fc\u015flerle ve i\u015flemlemeselci i\u015flevselcilikteki substrat\u0131n yoksulla\u015ft\u0131r\u0131lm\u0131\u015f do\u011fas\u0131yla keskin bir tezat olu\u015fturur. Ya\u015fam\u0131n maddesinde, i\u015flemlemesel bir kat\u0131l\u0131k de\u011fil, uyar\u0131labilir bir ak\u0131\u015fkanl\u0131k vard\u0131r.<\/strong><\/p>\n\n\n\n

Bu bak\u0131\u015f a\u00e7\u0131lar\u0131 \u00f6ng\u00f6r\u00fcc\u00fc i\u015fleme ve serbest enerji ilkesi anlat\u0131s\u0131n\u0131 tamamlar. Bili\u015f ve bilincin \u00f6zelliklerini ya\u015fam\u0131n \u00f6zellikleriyle ili\u015fkilendirirken biyolojik kuramlar\u0131n ba\u015fvurabilece\u011fi zengin kaynaklar\u0131 \u00f6rneklendirirler.<\/strong><\/p>\n\n\n\n

<\/div>\n\n\n\n

4.4 Ger\u00e7ek problem ve a\u00e7\u0131klama gedi\u011fi<\/em><\/p>\n\n\n\n

<\/div>\n\n\n\n

Biyolojik do\u011falc\u0131l\u0131\u011f\u0131 ikna edici bulmamam\u0131z\u0131n bir nedeni, bilincin do\u011fas\u0131na ili\u015fkin bir gizem duygusu tortusunu s\u00fcrd\u00fcrmesi ve bilinci do\u011frudan ya\u015fam\u0131n \u00f6zellikleriyle ili\u015fkilendirmenin, ba\u015flang\u0131\u00e7ta i\u015flevselci g\u00f6r\u00fc\u015fleri motive eden t\u00fcrden bir \u00f6zde\u015flik kuram\u0131 materyalizmiyle ters d\u00fc\u015fece\u011fine dair ku\u015fkunun buna e\u015flik etmesidir.<\/p>\n\n\n\n

Bu kayg\u0131n\u0131n, bilince dair tatmin edici bir bilimsel a\u00e7\u0131klaman\u0131n nas\u0131l g\u00f6r\u00fcnebilece\u011fini d\u00fc\u015f\u00fcnerek, en az\u0131ndan k\u0131smen etkisizle\u015ftirilebilece\u011fini d\u00fc\u015f\u00fcn\u00fcyorum. Bu yakla\u015f\u0131m, benim \u201cbilincin ger\u00e7ek problemi\u201d diye adland\u0131rd\u0131\u011f\u0131m \u015feye dayanmaktad\u0131r<\/strong> (Seth, 2016, 2021; ayr\u0131ca bkz. Searle, 2007). \u201cGer\u00e7ek problem\u201d yakla\u015f\u0131m\u0131, bilincin \u00f6zelliklerini alttaki d\u00fczenekler a\u00e7\u0131s\u0131ndan a\u00e7\u0131klama, \u00f6ng\u00f6rme<\/em> ve denetleme <\/em>yetene\u011fimiz artt\u0131k\u00e7a, bilincin maddeyle nas\u0131l ili\u015fkilendi\u011fine dair gizem duygusunun \u00e7\u00f6z\u00fclece\u011fini ve sonunda ortadan kalkabilece\u011fini ileri s\u00fcrer.<\/strong> Kusursuz olmaktan uzaksa da buradaki tarihsel analoji, (vitalist sezgilerde ifadesini bulan) ya\u015fam\u0131n kavramsal gizemlerinin nas\u0131l \u00e7\u00f6z\u00fcld\u00fc\u011f\u00fcn\u00fc hat\u0131rlat\u0131r. Biyologlar, ya\u015fayan sistemlerin \u00f6zelliklerini a\u00e7\u0131klama, \u00f6ng\u00f6rme ve denetleme konusunda daha yetkin h\u00e2le geldik\u00e7e s\u00f6z konusu gizemler de ortadan kalkm\u0131\u015ft\u0131r (Churchland, 1996; Seth, 2021; buna y\u00f6nelik bir itiraz i\u00e7in bkz. Chalmers, 2003).<\/p>\n\n\n\n

Bilincin \u00f6zellikleriyle ya\u015fam\u0131n \u00f6zellikleri aras\u0131ndaki ba\u011flar a\u00e7\u0131klay\u0131c\u0131 bir g\u00fc\u00e7 ta\u015f\u0131yorsa ve bunlar kaba indirgemeler de\u011filse, biyolojik do\u011falc\u0131l\u0131\u011f\u0131 daha fazla ciddiye almak bak\u0131m\u0131ndan elimiz rahatlar. Ya\u015f\u0131yor olman\u0131n \u00f6zellikleri, bilin\u00e7li olman\u0131n \u00f6zelliklerini a\u00e7\u0131kl\u0131yor, \u00f6ng\u00f6r\u00fcyor ve denetliyorsa, ikisi aras\u0131nda zorunlu bir ili\u015fkinin bulunmas\u0131 daha makul g\u00f6r\u00fcn\u00fcr.<\/strong><\/p>\n\n\n\n

<\/div>\n\n\n\n

4.5 \u00d6zet<\/em><\/p>\n\n\n\n

<\/div>\n\n\n\n

Ya\u015fam bilin\u00e7 i\u00e7in nas\u0131l bir \u00f6nem ta\u015f\u0131yor olabilir? Bilincin ancak ya\u015fayan sistemler olarak do\u011fam\u0131z \u0131\u015f\u0131\u011f\u0131nda anla\u015f\u0131labilece\u011fini ileri s\u00fcren epistemik bir g\u00f6r\u00fc\u015f mevcuttur.<\/strong> Bu g\u00f6r\u00fc\u015f\u00fc savunmak g\u00f6rece daha kolayd\u0131r, ancak a\u00e7\u0131klay\u0131c\u0131 bir \u00e7er\u00e7eve olarak \u00f6ng\u00f6r\u00fcc\u00fc i\u015flemenin kabul\u00fcne dayan\u0131r. \u00d6te yandan bir de ontolojik g\u00f6r\u00fc\u015f bulunmaktad\u0131r. Bu g\u00f6r\u00fc\u015fe g\u00f6re bilin\u00e7, daha g\u00fc\u00e7l\u00fc bir \u015fekilde biyolojik, ya\u015fayan substrata ba\u011f\u0131ml\u0131d\u0131r.<\/strong> Her iki g\u00f6r\u00fc\u015f de yaln\u0131zca alternatiflerinin (\u00f6rne\u011fin i\u015flemlemeselci i\u015flevselcili\u011fin) y\u00fczle\u015fti\u011fi sorunlar kar\u015f\u0131s\u0131nda bunlar\u0131 \u00f6ne s\u00fcrmek ya da neo-vitalist bir b\u00fcy\u00fcye geri \u00e7ekilmek yerine, o ya da bu \u015fekliyle biyolojik do\u011falc\u0131l\u0131k lehine olumlu arg\u00fcmanlar sunar.<\/p>\n\n\n\n

Bu olumlu arg\u00fcmanlar, genel olarak n\u00f6robili\u015fsel i\u015flevlerden ziyade \u00f6zellikle bilin\u00e7 i\u00e7in ge\u00e7erlidir. Bunun nedeni, biyolojik substrat\u0131n \u00f6zellikleriyle bilin\u00e7li deneyimin fenomenolojik \u00f6zellikleri aras\u0131ndaki ba\u011flant\u0131lar\u0131n, potansiyel olarak zengin bir a\u00e7\u0131klay\u0131c\u0131 de\u011fer ta\u015f\u0131mas\u0131d\u0131r.<\/strong> S\u0131n\u0131r durumunda, her bir bilin\u00e7li deneyimi, allostatik d\u00fczenlemeyi ve ya\u015famsal a\u00e7\u0131dan \u00f6nemli eylemleri koordine etme amac\u0131yla \u00e7ok kipli [multimodal<\/em>] bedensel ve \u00e7evresel sinyalleri (fenomenolojik olarak) b\u00fct\u00fcnle\u015ftiriyor olu\u015fuyla d\u00fc\u015f\u00fcnebiliriz <\/strong>(daha fazlas\u0131 i\u00e7in bkz. Seth, 2021; benzer bir g\u00f6r\u00fc\u015f i\u00e7in bkz. Kiverstein, 2020). Descartes\u2019\u0131n bir deyimini mezar\u0131ndan \u00e7\u0131kar\u0131p amac\u0131m\u0131za tabi k\u0131lacak olursak,<\/strong> bizler ya\u015fayan, etten kemikten \u201cmakine hayvanlar\u201d [beast machines<\/em>] oldu\u011fumuz i\u00e7in bilin\u00e7li varl\u0131klar\u0131z<\/strong> (Seth & Tsakiris, 2018).<\/strong><\/p>\n\n\n\n

Burada sunulan hi\u00e7bir \u015fey biyopsi\u015fizmi (t\u00fcm ya\u015fam\u0131n bilin\u00e7li oldu\u011fu g\u00f6r\u00fc\u015f\u00fcn\u00fc) desteklemez. Bu g\u00f6r\u00fc\u015f, bilin\u00e7li ya\u015fayan sistemler ile bilin\u00e7siz ya\u015fayan sistemler aras\u0131ndaki ayr\u0131m\u0131n nas\u0131l \u00e7izilece\u011fi gibi g\u00fc\u00e7 bir soruyu g\u00fcndeme getirir. Bu tart\u0131\u015fmay\u0131 burada ele almayaca\u011f\u0131m. Yaln\u0131zca t\u00fcm ya\u015fam bi\u00e7imlerinin, yukar\u0131da s\u00f6z\u00fc edilen \u00e7ok kipli ve ya\u015famsal a\u00e7\u0131dan \u00f6nemli b\u00fct\u00fcnle\u015ftirme t\u00fcr\u00fcne ihtiya\u00e7 duymayabilece\u011fini kaydetmekle yetinece\u011fim (yine de bkz. Lane, 2022). Ayn\u0131 \u015fekilde burada karbon \u015fovenizmini de (yaln\u0131zca karbon temelli ya\u015fam\u0131n bilin\u00e7li olabilece\u011fi iddias\u0131n\u0131) desteklemiyorum. Karbon temelli olmayan \u201cya\u015fayan\u201d sistemler de, \u00f6nemli oldu\u011funu savundu\u011fum t\u00fcrden substrata ba\u011f\u0131ml\u0131 \u00e7ok \u00f6l\u00e7ekli etkinli\u011fe dahil olabilir. Ba\u015fka bir deyi\u015fle, ilgili s\u00fcre\u00e7lerin karbonun \u00f6tesinde de ger\u00e7ekle\u015ftirilmesi i\u00e7in yeterli substrat esnekli\u011fine sahip olabilir. Bu durum, neyin \u201cya\u015fayan sistem\u201d say\u0131laca\u011f\u0131na dair ba\u015fka g\u00fc\u00e7 sorular\u0131 g\u00fcndeme getirir. Bunlar da ba\u015fka bir g\u00fcn\u00fcn sorular\u0131d\u0131r (Ball, 2023; Lane, 2015; Schr\u00f6dinger, 1944).<\/p>\n\n\n\n

“Bizler ya\u015fayan, etten kemikten \u201cmakine hayvanlar\u201d [beast machines<\/em>] oldu\u011fumuz i\u00e7in bilin\u00e7li varl\u0131klar\u0131z.”<\/p><\/blockquote><\/figure>\n\n\n\n

<\/div>\n\n\n\n

5.0 Ger\u00e7ek yapay bilin\u00e7 senaryolar\u0131<\/strong><\/p>\n\n\n\n

<\/div>\n\n\n\n

Bilin\u00e7li YZ\u2019ye kar\u015f\u0131 sav\u0131m iki k\u0131s\u0131mdan olu\u015fuyor. Olumsuz k\u0131s\u0131m, insanlar\u0131 bilin\u00e7li YZ\u2019nin ka\u00e7\u0131n\u0131lmaz ya da m\u00fcmk\u00fcn oldu\u011funu d\u00fc\u015f\u00fcnmeye g\u00f6t\u00fcren varsay\u0131mlar\u0131 sorgulamaktad\u0131r.<\/strong> \u015eeylere temellendirilmemi\u015f bilin\u00e7 atf\u0131na [over-attribution<\/em>] yol a\u00e7abilecek \u00f6nyarg\u0131lar kaydedilmekte ve bilin\u00e7li YZ\u2019nin dayand\u0131\u011f\u0131 i\u015flemlemeselci i\u015flevselcilik ile silikon substrat esnekli\u011fi varsay\u0131mlar\u0131na itiraz edilmektedir. Olumlu k\u0131s\u0131m, aktif \u00e7\u0131kar\u0131m, sibernetik, otopoiesis ve serbest enerji ilkesine dayanan bir t\u00fcr biyolojik do\u011falc\u0131l\u0131k lehine savlar ileri s\u00fcrmektedir. Olumsuz arg\u00fcmanlar do\u011fru yoldaysa, YZ\u2019nin halihaz\u0131rdaki substrattan ba\u011f\u0131ms\u0131z do\u011frultular\u0131 i\u00e7in bilin\u00e7li YZ m\u00fcmk\u00fcn olmayacakt\u0131r. Olumlu arg\u00fcmanlar do\u011fru yoldaysa, ger\u00e7ek yapay bilin\u00e7 ancak ilgili anlamda canl\u0131 olan makineler yarat\u0131l\u0131rsa m\u00fcmk\u00fcn olacakt\u0131r.<\/strong><\/p>\n\n\n\n

Her iki k\u0131sm\u0131 da buraya kadar ortaya koydu\u011fumuza g\u00f6re, bilin\u00e7li YZ\u2019nin ortaya \u00e7\u0131kabilece\u011fi \u00e7e\u015fitli senaryolar\u0131 \u00f6zetlemenin zaman\u0131 geldi. Bug\u00fcn\u00fcn yapay zek\u00e2s\u0131ndan basit ekstrapolasyonlardan ba\u015flay\u0131p biyolojik do\u011falc\u0131l\u0131\u011f\u0131n g\u00fc\u00e7l\u00fc bi\u00e7imlerine uzanan a\u015famalar halinde ve m\u00fcmk\u00fcn olduk\u00e7a ilgili bilin\u00e7 kuramlar\u0131na at\u0131flarla bu \u00f6zeti sunaca\u011f\u0131m (Tablo 1). Bir kuramdan s\u00f6z etti\u011fimde, s\u00f6z konusu kurama dayanan bilin\u00e7li YZ\u2019nin bir ko\u015fulu, bilin\u00e7 i\u00e7in ger\u00e7ekten yeterli ko\u015fullar\u0131 a\u00e7\u0131k\u00e7a belirtmesidir. Bu \u00e7\u0131tay\u0131 y\u00fckse\u011fe koyar. Zira yaln\u0131zca mevcut kuramlar\u0131n t\u00fcm\u00fcn\u00fcn muhtemelen yanl\u0131\u015f ya da en iyi ihtimalle eksik olmas\u0131 de\u011fil, ayn\u0131 zamanda bir\u00e7o\u011funun genel olarak yeterli ko\u015fullar \u00f6ne s\u00fcrmek yerine bilincin var oldu\u011funa inan\u0131lan yerlerde olas\u0131 d\u00fczenekleri betimlemekle s\u0131n\u0131rl\u0131 kalmas\u0131 s\u00f6z konusudur<\/strong> (Butlin ve di\u011ferleri, 2023; Kuhn, 2024; Mudrik ve di\u011ferleri, 2025; Seth & Bayne, 2022).<\/strong><\/p>\n\n\n\n

SENARYO<\/strong><\/td>A\u00c7IKLAMA<\/strong><\/td>\u0130\u015eLEMLEMESELC\u0130 \u0130\u015eLEVSELC\u0130L\u0130\u011e\u0130 VARSAYAR MI?<\/strong><\/td>SUBSTRAT (TEMEL MALZEME) \u00d6NEML\u0130 M\u0130?<\/strong><\/td>\u00d6RNEK<\/strong><\/td><\/tr><\/thead>
Naif\u00e7e Gidi\u015fata B\u0131rakma<\/strong><\/td>Yapay zek\u00e2 ak\u0131lland\u0131k\u00e7a bilin\u00e7 de kendili\u011finden ortaya \u00e7\u0131kacakt\u0131r<\/td>Evet (Turing)<\/td>Hay\u0131r<\/td>B\u00fcy\u00fck dil modelleri (LLM’ler)<\/td><\/tr>
Kuram Temelli \u0130\u015flemlemeselci<\/strong><\/td>Bilin\u00e7, i\u015flemlemeselci bilin\u00e7 kuramlar\u0131 uyguland\u0131\u011f\u0131nda ortaya \u00e7\u0131kar<\/td>Evet (Turing)<\/td>Hay\u0131r<\/td>Dikkat \u015femas\u0131 kuram\u0131\u00b9, k\u00fcresel \u00e7al\u0131\u015fma alan\u0131 kuram\u01312<\/sup>, (baz\u0131) \u00fcst d\u00fczey d\u00fc\u015f\u00fcnce kuramlar\u0131\u00b3<\/td><\/tr>
Substrata Ba\u011fl\u0131 \u0130\u015flemlemeselci<\/strong><\/td>Bilin\u00e7, yaln\u0131zca belirli substratlarda uygulanabilen i\u015flemlemelere ba\u011fl\u0131d\u0131r<\/td>Evet (potansiyel olarak Turing’den daha geni\u015f)<\/td>Evet, ilgili i\u015flemlemeleri uygulamak i\u00e7in<\/td>\u00d6l\u00fcml\u00fc i\u015flemleme\u2074, sinirbi\u00e7imsel i\u015flemleme, sinirsel i\u015flemleme\u2075, biyolojik i\u015flemleme\u2076<\/td><\/tr>
Substrata Ba\u011fl\u0131 (Zay\u0131f)<\/strong><\/td>Bilin\u00e7, i\u015flemlemesel olmayan i\u015flevsel \u00f6rg\u00fctlenmeye ba\u011fl\u0131d\u0131r<\/td>Hay\u0131r<\/td>Evet, ilgili i\u015flevsel \u00f6rg\u00fctlenmeyi uygulamak i\u00e7in<\/td>\u0130\u015flemlemeselci olmayan sinirbi\u00e7imsel yakla\u015f\u0131mlar, dinamik kuramlar\u0131n uygulamalar\u0131, (IIT – i\u015flev i\u00e7in de\u011fil, neden-sonu\u00e7 yap\u0131s\u0131 i\u00e7in\u2077)<\/td><\/tr>
Substrata Ba\u011fl\u0131 (G\u00fc\u00e7l\u00fc)<\/strong><\/td>Bilin\u00e7, biyolojik temelinin (g\u00f6r\u00fcn\u00fc\u015fte) i\u00e7kin \u00f6zelliklerine ba\u011fl\u0131d\u0131r<\/td>Hay\u0131r<\/td>Evet, substrata \u00f6zg\u00fc i\u015flevleri sa\u011flamak i\u00e7in veya (g\u00f6r\u00fcn\u00fc\u015fte) i\u00e7kin \u00f6zellikler sayesinde<\/td>Serebral organoidler, hibrit sistemler\u2078, sentetik biyoloji<\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n

Tablo 1. Bilin\u00e7li YZ i\u00e7in senaryolar. “\u0130\u015flemleme\u201dnin kapsam\u0131 Turing\u2019in tan\u0131m\u0131n\u0131n \u00f6tesine geni\u015fletilirse, 3 ve 4 numaral\u0131 senaryolar aras\u0131ndaki ayr\u0131m bulan\u0131kla\u015fabilir ve 3 numaral\u0131 senaryoya daha \u00e7ok olas\u0131l\u0131k s\u0131\u011fd\u0131r\u0131labilir. Biyolojik do\u011falc\u0131l\u0131k 4 ve 5 numaral\u0131 senaryolar alt\u0131na d\u00fc\u015fer ve \u201ci\u015flemlemeselci biyolojik do\u011falc\u0131l\u0131k\u201d i\u00e7in 3 numaral\u0131 senaryoya da potansiyel olarak girebilir. 1<\/sup>(Graziano, 2017), 2<\/sup>(VanRullen & Kanai, 2021), 3<\/sup>(Dehaene ve di\u011ferleri, 2017), 4<\/sup>(Ororbia & Friston, 2023), 5<\/sup>(Piccinini, 2020), 6<\/sup>(Kirkpatrick, 2022), 7<\/sup>(Findlay ve di\u011ferleri, 2024), IIT\u2019nin bilince ili\u015fkin olarak i\u015flevsel \u00f6rg\u00fctlenmeden de\u011fil neden-sonu\u00e7 yap\u0131s\u0131ndan yana tav\u0131r ald\u0131\u011f\u0131na dikkat ediniz 8<\/sup>(Morales Pantoja ve di\u011ferleri, 2023).<\/em><\/p>\n\n\n\n

<\/div>\n\n\n\n

5.1 Naif\u00e7e gidi\u015fata b\u0131rakma<\/em><\/p>\n\n\n\n

<\/div>\n\n\n\n

Bilin\u00e7 bir i\u015flemleme bi\u00e7imidir ve YZ ak\u0131lland\u0131k\u00e7a<\/em>, belki de Yapay Genel Zek\u00e2 (YGZ) e\u015fi\u011finde, kendili\u011finden ortaya \u00e7\u0131kacakt\u0131r<\/em>. <\/strong>Bu senaryonun b\u00fcy\u00fck \u00f6l\u00e7\u00fcde psikolojik \u00f6nyarg\u0131lara ve i\u015flemlemeselci i\u015flevselcilik hakk\u0131ndaki yeterince s\u0131nanmam\u0131\u015f varsay\u0131mlara dayand\u0131\u011f\u0131n\u0131 ve bu nedenle ikna edici olmad\u0131\u011f\u0131n\u0131 savunuyorum.<\/p>\n\n\n\n

<\/div>\n\n\n\n

5.2 Kuram temelli i\u015flemlemeselci senaryo<\/em><\/p>\n\n\n\n

<\/div>\n\n\n\n

Bilin\u00e7 bir i\u015flemleme bi\u00e7imidir, ancak i\u015flemlemeselci bilin\u00e7 kuramlar\u0131na dayanarak a\u00e7\u0131k\u00e7a tasarlanmas\u0131 [designed in] gerekmektedir<\/em>.<\/strong> Bu senaryoyla ilgili kuramlar aras\u0131nda, bilincin \u00e7ok kipli k\u00fcresel bir sinirsel \u00e7al\u0131\u015fma alan\u0131 i\u00e7inde enformasyon payla\u015f\u0131m\u0131na ba\u011fl\u0131 oldu\u011funu \u00f6ne s\u00fcren k\u00fcresel \u00e7al\u0131\u015fma alan\u0131 kuram\u0131 [global workspace theory<\/em>] (Mashour ve di\u011ferleri, 2020); bilincin, alt-d\u00fczey zihinsel durumlar\u0131n i\u00e7eri\u011fini bir t\u00fcr \u201cbilin\u00e7li olma\u201d olarak temsil eden \u00fcst d\u00fczey zihinsel durumlara ba\u011fl\u0131 oldu\u011funu savunan \u00fcst d\u00fczey [higher-order<\/em>] d\u00fc\u015f\u00fcnce kuramlar\u0131 (Brown ve di\u011ferleri, 2019; Fleming, 2020) ve bilincin, dikkat denetiminin modellenmesi sonucu olarak ortaya \u00e7\u0131kt\u0131\u011f\u0131n\u0131 ileri s\u00fcren dikkat \u015femas\u0131 [attention schema<\/em>] kuram\u0131 yer al\u0131r (Graziano, 2017). Bu senaryoda bilin\u00e7li YZ\u2019nin ikna edicili\u011fi yine i\u015flemlemeselci i\u015flevselcili\u011fin do\u011fru olmas\u0131na ve silikon substrat esnekli\u011fine ba\u011fl\u0131d\u0131r.<\/strong> Butlin ve meslekta\u015flar\u0131 (2023) bu senaryonun kapsaml\u0131 bir \u00f6zetini sunup s\u00f6z konusu kuramlar\u0131n ilkelerinin mevcut YZ mimarilerinde ne \u00f6l\u00e7\u00fcde \u00f6rt\u00fck olarak bulunabilece\u011fini incelerken, yak\u0131n zamanda Dung ve Kersten (2024) i\u015flemlemeselci g\u00f6r\u00fc\u015fleri savunmu\u015ftur.<\/p>\n\n\n\n

<\/div>\n\n\n\n

5.3 Substrata ba\u011f\u0131ml\u0131 i\u015flemlemeselci<\/em> senaryo<\/em><\/p>\n\n\n\n

<\/div>\n\n\n\n

Bilin\u00e7 i\u015flemlemeseldir, fakat ilgili i\u015flemlemeler yaln\u0131zca belirli substratlar taraf\u0131ndan uygulanabilir<\/em>.<\/strong> Burada, bilincin \u00f6l\u00fcml\u00fc i\u015flemleme ile ilgili bir mesele olmas\u0131, yani bilin\u00e7le ili\u015fkili i\u015flemlemelerin substrat\u0131ndan ayr\u0131lamaz olmas\u0131 s\u00f6z konusu olabilir (B\u00f6l\u00fcm 3.4). Daha genel olarak, bilincin alt\u0131nda yatan i\u015flemlemelerin yaln\u0131zca belirli substratlarda uygulanabilmesinin do\u011fa yasalar\u0131yla ilgili bir mesele olmas\u0131 m\u00fcmk\u00fcnd\u00fcr.<\/strong> Bu g\u00f6r\u00fc\u015f, substrata daha fazla ba\u011f\u0131ml\u0131l\u0131\u011fa izin veren daha geni\u015f \u201ci\u015flemleme\u201d yorumlar\u0131 i\u00e7in de daha makul g\u00f6r\u00fcn\u00fcr (N. G. Anderson & G. Piccinini, 2024; Kirkpatrick, 2022). Bu senaryonun bir g\u00fc\u00e7l\u00fc\u011f\u00fc, zihin ve bilin\u00e7 hakk\u0131nda i\u015flemlemeselci g\u00f6r\u00fc\u015fleri \u00e7ekici k\u0131lan g\u00fcd\u00fclerden birinin (elbette yeg\u00e2ne de\u011fil), substrat ba\u011f\u0131ms\u0131zl\u0131\u011f\u0131 fikrinden sezgisel olarak \u00e7\u0131kan geni\u015f kapsaml\u0131 \u00e7oklu ger\u00e7ekle\u015febilirlik olmas\u0131d\u0131r. Bildi\u011fim kadar\u0131yla bu senaryodan istifade eden bir bilin\u00e7 kuram\u0131 bulunmamaktad\u0131r.<\/p>\n\n\n\n

Bu senaryo, \u201ci\u015flemlemeselci biyolojik do\u011falc\u0131l\u0131k\u201d diyebilece\u011fimiz, yaln\u0131zca ya\u015fayan sistemlerin ilgili i\u015flemlemeleri uygulayabilmesi gibi ilgin\u00e7 bir olas\u0131l\u0131\u011f\u0131 da i\u00e7erir.[37]<\/a> Bu g\u00f6r\u00fc\u015f\u00fc destekleyen bir motivasyon, evrimin, \u00f6rne\u011fin enerji verimlili\u011fi, bozulma [degeneracy<\/em>] ve sa\u011flaml\u0131k [robustness<\/em>] gibi k\u0131s\u0131tlar alt\u0131nda, i\u015flemlemeyi g\u00fc\u00e7l\u00fc bir i\u015flevsel \u00e7\u00f6z\u00fcm olarak se\u00e7mi\u015f olmas\u0131 ihtimalidir. Bu durumda, substrata ba\u011f\u0131ml\u0131 (biyolojik) i\u015flemleme bi\u00e7imleri, bilin\u00e7 kuramlar\u0131 i\u00e7in daha \u00e7ekici bir zemin haline gelebilir. Buradaki bir g\u00fc\u00e7l\u00fck de \u015fu ki, hangi t\u00fcrden i\u015flemlemelerin (\u00f6rne\u011fin \u00f6l\u00fcml\u00fc, n\u00f6ral, kuantum-sonras\u0131 s\u00fcper-Turing vb.) yaln\u0131zca ya\u015fayan sistemlerde uygulanabilece\u011fini ve bunun neden b\u00f6yle oldu\u011funu a\u00e7\u0131klamak gerekir.<\/p>\n\n\n\n

Bu genel senaryoda, bilin\u00e7li YZ\u2019ye, yeterli (muhtemelen \u00f6l\u00fcml\u00fc) i\u015flemlemeleri uygulayabilen sinirbi\u00e7imsel i\u015flemleme bi\u00e7imleriyle ula\u015f\u0131lmas\u0131 ihtimal dahilindedir<\/strong> (E\u011fer i\u015flemlemeselci<\/strong> biyolojik do\u011falc\u0131l\u0131k ge\u00e7erliyse ya\u015fam da zorunlu olacakt\u0131r).<\/strong> Standart dijital i\u015flemlemenin s\u00f6z konusu i\u015flemlemelere yakla\u015fabilece\u011fi d\u00fc\u015f\u00fcn\u00fclebilir, fakat bu kestirimlerin bilinci ger\u00e7ekle\u015ftirmeye yetecek kadar iyi olup olmayaca\u011f\u0131 ucu a\u00e7\u0131k bir sorudur. Ayr\u0131ca, sim\u00fclasyonda uygulanan i\u015flemlemelerin sim\u00fcle edilen i\u015flemlemelerden farkl\u0131 olmas\u0131n\u0131n sonucu de\u011fi\u015ftirip de\u011fi\u015ftirmeyece\u011fine dair sorular\u0131n da ucu a\u00e7\u0131kt\u0131r.[38]<\/a><\/p>\n\n\n\n

<\/div>\n\n\n\n

5.4 Substrata ba\u011f\u0131ml\u0131, zay\u0131f senaryo<\/em><\/p>\n\n\n\n

<\/div>\n\n\n\n

Bilin\u00e7, substrata ba\u011f\u0131ml\u0131 i\u015flemlemesel olmayan i\u015flevsel \u00f6zelliklere dayan\u0131r.<\/em> Bu \u00f6zelliklere, devam eden s\u00fcre\u00e7ler, alan etkileri, stokastik etkiler, ince taneli zamanlama ili\u015fkileri ve n\u00f6robiyolojik sistemlerin di\u011fer (i\u015flemlemesel olmayan) i\u015flevsel \u00f6zellikleri d\u00e2hil edilebilir. Bu senaryo, yaln\u0131zca belirli substratlar taraf\u0131ndan uygulanabilen i\u015flevsel \u00f6zellikleri kapsar, fakat metabolizma, sindirim ve benzeri gibi belirli bir maddeye ya da substrata g\u00f6re tan\u0131mlanan<\/em> i\u015flevleri d\u0131\u015far\u0131da tutar. Substrat ile i\u015flev aras\u0131nda h\u00e2l\u00e2 net bir ayr\u0131m vard\u0131r.<\/strong><\/p>\n\n\n\n

Burada yaln\u0131zca ya\u015fayan sistemlerin bilin\u00e7 i\u00e7in gerekli i\u015flevsel \u00f6rg\u00fctlenmeyi uygulayabilece\u011fi gibi \u00f6nemli bir olas\u0131l\u0131kla kar\u015f\u0131la\u015f\u0131r\u0131z. Bu, 4. b\u00f6l\u00fcmde \u00f6zetlenen \u00f6ng\u00f6r\u00fcc\u00fc i\u015fleme ve serbest enerji ilkesinin dinamik yorumlar\u0131yla uyumlu, \u201czay\u0131f\u201d bir biyolojik do\u011falc\u0131l\u0131k olacakt\u0131r. Bu olanak, yukar\u0131da s\u00f6z\u00fc edilen \u201ci\u015flemlemeselci biyolojik do\u011falc\u0131l\u0131k\u201dtan daha cazip g\u00f6r\u00fcn\u00fcr, \u00e7\u00fcnk\u00fc substrat \u00f6zellikleri bu g\u00f6r\u00fc\u015f\u00fc k\u00f6steklemek yerine destekleyecektir. 3.6. b\u00f6l\u00fcmde bahsedildi\u011fi gibi \u00e7e\u015fitli dinamik bilin\u00e7 kuramlar\u0131 da bu senaryonun i\u00e7ine girer. Elektromanyetik alanlar\u0131n bilin\u00e7 durumlar\u0131 i\u00e7in fiziksel dayanak sa\u011flad\u0131\u011f\u0131n\u0131 \u00f6ne s\u00fcren elektromanyetik alan kuramlar\u0131 da ayn\u0131 \u015fekilde bu kapsamdad\u0131r [(McFadden, 2020), ayr\u0131ca bkz. (Lane, 2022)].<\/p>\n\n\n\n

Bu senaryoda bilin\u00e7 bir i\u015flemleme meselesi de\u011fildir ve bilin\u00e7li YZ\u2019nin ikna edicili\u011fi, ilgili i\u015flevsel \u00f6zelliklerin alternatif bir substratta uygulan\u0131p uygulanamayaca\u011f\u0131na ba\u011fl\u0131d\u0131r. Substrat esnekli\u011finden s\u00f6z etmek, her ne kadar bu senaryoda m\u00fcmk\u00fcn olsa da, i\u015flemlemenin ilkece sundu\u011fu substrat ba\u011f\u0131ms\u0131zl\u0131\u011f\u0131ndan art\u0131k yararlan\u0131lamad\u0131\u011f\u0131 i\u00e7in zor g\u00f6r\u00fcn\u00fcr.<\/strong> Bu \u00f6zelliklerin i\u015flemlemesel sim\u00fclasyonlar\u0131n\u0131n bilinci ger\u00e7ekle\u015ftirmesi art\u0131k g\u00fcvence alt\u0131nda de\u011fildir. Bu ancak i\u015flemlemeselci i\u015flevselcilik zaten do\u011fruysa ve sim\u00fclasyonlar yeterince iyiyse s\u00f6z konusu olabilir (ki bu da, canl\u0131 sistemlerin ayr\u0131nt\u0131l\u0131 ve fazlas\u0131yla optimize edilmi\u015f maddi \u00f6zelliklerinin \u00f6nemli oldu\u011fu ortaya \u00e7\u0131karsa a\u015f\u0131r\u0131 derecede zor olacakt\u0131r).<\/p>\n\n\n\n

Elimizdeki ve bir \u00f6nceki senaryo aras\u0131ndaki s\u0131n\u0131r, neyin \u201ci\u015flemleme\u201d say\u0131laca\u011f\u0131na ba\u011fl\u0131d\u0131r. 3.4. b\u00f6l\u00fcmdeki gibi daha geni\u015f i\u015flemleme anlay\u0131\u015flar\u0131na m\u00fcsaade edilirse, kimi i\u015flemlemesel olmayan i\u015flevler i\u015flemlemesel say\u0131lacakt\u0131r. Fakat bilin\u00e7li YZ\u2019ye ili\u015fkin ihtimaller pek de\u011fi\u015fmeyecektir, \u00e7\u00fcnk\u00fc bu daha geni\u015f i\u015flemleme anlay\u0131\u015flar\u0131 b\u00fcy\u00fck ihtimalle gerekli substrat esnekli\u011finden yoksundur.<\/p>\n\n\n\n

<\/div>\n\n\n\n

5.5 Substrata ba\u011f\u0131ml\u0131, g\u00fc\u00e7l\u00fc senaryo<\/em><\/p>\n\n\n\n

<\/div>\n\n\n\n

Bilin\u00e7, maddi biyolojik temelinin i\u00e7kin \u00f6zelliklerine ba\u011fl\u0131d\u0131r<\/em>. Bu senaryo, bilincin, substrat\u0131n\u0131n maddeselli\u011fine i\u00e7kin oldu\u011fu sezgisini yans\u0131t\u0131r.<\/strong> Bunu yorumlaman\u0131n (en az) iki yolu vard\u0131r. \u0130lki, i\u015flevsel yorumdur, bilin\u00e7 i\u00e7in gerekli i\u015flevler tan\u0131m gere\u011fi substrat\u0131n baz\u0131 y\u00f6nlerini de i\u00e7erir (\u00f6rne\u011fin sindirim i\u015flevi zorunlu olarak ger\u00e7ek g\u0131day\u0131 i\u00e7erir). \u0130kincisi, i\u015flevlerin etraf\u0131ndan dolan\u0131r ve bilinci, Searle\u2019\u00fcn \u00f6zg\u00fcn biyolojik do\u011falc\u0131l\u0131k kavray\u0131\u015f\u0131n\u0131n ruhuna uygun bi\u00e7imde, do\u011frudan substrat\u0131n\u0131n i\u00e7kin \u00f6zelliklerine ba\u011flar. Searle\u2019\u00fc izleyerek ben de konuyla alakal\u0131 substrat olarak ya\u015fayan maddeye odaklan\u0131yorum.<\/p>\n\n\n\n

Biyolojik do\u011falc\u0131l\u0131k lehindeki sav\u0131m bu senaryoya y\u00f6nelmektedir.<\/strong> \u00d6ng\u00f6r\u00fcc\u00fc i\u015flemeden ve denetimli hal\u00fcsinasyonlardan serbest enerji ilkesine, metabolizmaya ve otopoiesise uzanan hat, biraz da \u201cger\u00e7ek problem\u201d kar\u015f\u0131s\u0131ndaki tevazuyla birlikte tatland\u0131r\u0131ld\u0131\u011f\u0131nda, bilincin biyolojiye g\u00fc\u00e7l\u00fc bir ba\u011f\u0131ml\u0131l\u0131\u011f\u0131na i\u015faret eder. G\u00f6rebildi\u011fim kadar\u0131yla bu yakla\u015f\u0131m, yukar\u0131da verilen iki yorum aras\u0131nda agnostiktir, \u00e7\u00fcnk\u00fc bir d\u00fczeyde i\u00e7kin g\u00f6r\u00fcnen \u00f6zellikler, her zaman daha ince taneli d\u00fczeylerdeki i\u015flevsel \u00f6rg\u00fctlenme \u00f6r\u00fcnt\u00fcleri a\u00e7\u0131s\u0131ndan anla\u015f\u0131labilir. Ama e\u011fer ya\u015fayan sistemlerin i\u00e7kin oldu\u011fu varsay\u0131lan metabolik, uyar\u0131labilir ve otopoietik \u00f6zellikleri, di\u011fer maddi bile\u015fenlerin ve s\u00fcre\u00e7lerin daha ince taneli i\u015flevsel betimlemeleriyle temellendirilebiliyorsa vars\u0131n \u00f6yle olsun. Ya\u015fam yine de \u00f6nem ta\u015f\u0131yacakt\u0131r ve bunu \u201cya\u015fam\u201d dememizi hakl\u0131 \u00e7\u0131karan \u00f6zellikleri sayesinde yapacakt\u0131r. Her iki durumda da, substrat ile i\u015flev aras\u0131nda bir b\u00f6l\u00fcnme yerine art\u0131k derin bir s\u00fcreklilik s\u00f6z konusudur. Bu senaryoda bilin\u00e7 bir i\u015flemleme meselesi de\u011fildir ve bilin\u00e7li YZ\u2019nin \u201cya\u015fayan\u201d bir YZ olmas\u0131 gerekecektir.<\/strong>[39]<\/a><\/p>\n\n\n\n

“Ya\u015fam yine de \u00f6nem ta\u015f\u0131yacakt\u0131r ve bunu \u201cya\u015fam\u201d dememizi hakl\u0131 \u00e7\u0131karan \u00f6zellikleri sayesinde yapacakt\u0131r.”<\/p><\/blockquote><\/figure>\n\n\n\n

<\/div>\n\n\n\n

5.6 Di\u011fer kuramlar<\/em><\/p>\n\n\n\n

<\/div>\n\n\n\n

\u0130\u015flemlemeselci i\u015flevselcilik ile biyolojik do\u011falc\u0131l\u0131k aras\u0131ndaki yelpazeye odakland\u0131\u011f\u0131mdan bu iki u\u00e7 \u00fczerinde do\u011fal bi\u00e7imde yer almayan bilin\u00e7 kuramlar\u0131n\u0131 \u015fu ana dek ihmal ettim. Ancak bu kuramlar\u0131n yeterince do\u011fru yolda oldu\u011fu ortaya \u00e7\u0131karsa, onlar\u0131n da bilin\u00e7li YZ\u2019nin ikna edicili\u011fi a\u00e7\u0131s\u0131ndan sonu\u00e7lar\u0131 olacakt\u0131r. Bu y\u00fczden bu kuramlar\u0131 k\u0131saca anmakta fayda vard\u0131r.<\/p>\n\n\n\n

B\u00fct\u00fcnle\u015fik enformasyon kuram\u0131 [integrated information theory, IIT<\/em>] (Albantakis ve di\u011ferleri, 2023; Tononi ve di\u011ferleri, 2016) bilin\u00e7 i\u00e7in yeterli ko\u015fullar\u0131 a\u00e7\u0131k\u00e7a dile getirmesi ve yo\u011fun tart\u0131\u015fmalar\u0131n oda\u011f\u0131 olmas\u0131yla \u00f6ne \u00e7\u0131km\u0131\u015ft\u0131r (Gomez-Marin & Seth, 2025). IIT\u2019ye g\u00f6re bilin\u00e7li bir deneyim, indirgenemez b\u00fct\u00fcnle\u015fik bilginin azami oldu\u011fu bir fiziksel substrattan ortaya \u00e7\u0131kan neden-sonu\u00e7 yap\u0131s\u0131yla \u00f6zde\u015ftir. Bu t\u00fcr azamilerin [maxima<\/em>] bulundu\u011fu her yerde bilin\u00e7 vard\u0131r. Bu kurama g\u00f6re ger\u00e7ek yapay bilin\u00e7, yaln\u0131zca do\u011fru t\u00fcrden i\u00e7sel nedensel yap\u0131ya sahip sistemler i\u00e7in m\u00fcmk\u00fcn olacakt\u0131r<\/strong> (Findlay ve di\u011ferleri, 2024). Salt ileri beslemeli bir sinir a\u011f\u0131 bilin\u00e7 i\u00e7in s\u0131f\u0131r de\u011ferine sahip olacakt\u0131r; buna kar\u015f\u0131l\u0131k, do\u011fru t\u00fcrden ve fiziksel olarak ger\u00e7ekle\u015ftirilmi\u015f yinelemeli ba\u011flant\u0131sall\u0131\u011fa sahip bir sinir a\u011f\u0131, neyden yap\u0131lm\u0131\u015f oldu\u011fundan ba\u011f\u0131ms\u0131z olarak, s\u0131f\u0131rdan b\u00fcy\u00fck bir bilin\u00e7 d\u00fczeyine sahip olacakt\u0131r. IIT do\u011fruysa, geleneksel YZ bilin\u00e7 yolunda de\u011fildir. Fakat ya\u015fam\u0131n da bu a\u00e7\u0131dan \u00f6zel bir yan\u0131 oldu\u011fu s\u00f6ylenemez.<\/strong><\/p>\n\n\n\n

Kuantum mekani\u011fine dayal\u0131 bilin\u00e7 kuramlar\u0131 (\u00d6r. Hameroff & Penrose, 2014; Neven ve di\u011ferleri, 2024) uzun s\u00fcredir tart\u0131\u015fmal\u0131d\u0131r (Koch & Hepp, 2006; Seth, 2012). Bununla birlikte kuantum mekani\u011fi son derece g\u00fc\u00e7l\u00fc bir fizik kuram\u0131d\u0131r ve kuantum etkilerinin \u00e7e\u015fitli biyolojik s\u00fcre\u00e7lerde \u00f6nemli oldu\u011fu bilinmektedir (McFadden & Al-Khalili, 2018; Mohseni ve di\u011ferleri, 2013). Bilin\u00e7le ili\u015fkili biyolojik s\u00fcre\u00e7lerde, \u00f6rne\u011fin anestezide, kuantum etkilerinin rol oynayabilmesi m\u00fcmk\u00fcnd\u00fcr (Kalra ve di\u011ferleri, 2023; Li ve di\u011ferleri, 2018). Ancak b\u00f6yle olsa bile bu durum, bilincin genel olarak zorunlu bir \u00f6zelli\u011fi de\u011fil, biyolojik bilince \u00f6zg\u00fc olumsal bir \u00f6zellik olabilir ve yapay bilin\u00e7 a\u00e7\u0131s\u0131ndan bir sonu\u00e7 do\u011furmayabilir. Kuantum etkilerinin bilin\u00e7te daha temel bir rol oynad\u0131\u011f\u0131 ortaya \u00e7\u0131karsa <\/strong>(Hameroff & Penrose, 2014), ger\u00e7ek yapay bilincin gerekli kuantum-d\u00fczeyi etkileri \u00f6rneklemesi gerekece\u011finden, geleneksel YZ devre d\u0131\u015f\u0131 kalacakt\u0131r. Buna yak\u0131n bir olas\u0131l\u0131k da bilincin kuantum t\u00fcr\u00fc i\u015flemlemelere ba\u011fl\u0131 olmas\u0131d\u0131r. <\/strong>\u201cKuantum i\u015flemlemeselci i\u015flevselcilikte\u201d bilin\u00e7 yaln\u0131zca, kuantum bilgisayarlar taraf\u0131ndan ger\u00e7ekle\u015ftirilebilir, geleneksel bilgisayarlar taraf\u0131ndan de\u011fil (Neven ve di\u011ferleri, 2024). Bu senaryolar\u0131n ikna edicili\u011fi, s\u00f6z konusu kuantum kuramlar\u0131n\u0131n ne kadar ciddiye al\u0131nd\u0131\u011f\u0131na ba\u011fl\u0131d\u0131r.<\/p>\n\n\n\n

<\/div>\n\n\n\n

5.7 Bedenlenmi\u015flik ve g\u00f6m\u00fcl\u00fcl\u00fck<\/em><\/p>\n\n\n\n

<\/div>\n\n\n\n

Biyolojik beyinler bedenler i\u00e7inde evrilmi\u015f, geli\u015fmi\u015f ve i\u015fler h\u00e2le gelmi\u015ftir. Bu bedenler de \u00e7evrelerin i\u00e7ine g\u00f6m\u00fcl\u00fcd\u00fcr. YZ sistemleri ise tipik olarak bedenden yoksundur ve \u201c\u00e7evre\u201dleriyle yaln\u0131zca pasif bi\u00e7imde kar\u015f\u0131la\u015fabilirler.<\/strong> LLM\u2019ler iyi bir \u00f6rnektir. Ger\u00e7i bu sistemler gitgide, sanal \u00e7evrelerle etkile\u015fime girdikleri \u201ceyleyici\u201d ba\u011flamlarda konu\u015fland\u0131r\u0131lmaktad\u0131r (\u00f6rne\u011fin yaz\u0131l\u0131m i\u00e7inde g\u00f6revler icra ederek). Baz\u0131 bilin\u00e7 kuramlar\u0131 bedenlenmi\u015fli\u011fi ve g\u00f6m\u00fcl\u00fcl\u00fc\u011f\u00fc a\u00e7\u0131k\u00e7a gerekli sayar (\u00f6rne\u011fin Damasio, 2000; Edelman, 1989; O’Regan & No\u00eb, 2001; Seth, 2021; Solms, 2021) ve bu \u00f6zellikler uzun zamand\u0131r bili\u015fsel bilime ili\u015fkin enaktif perspektiflerde vurgulanmaktad\u0131r (bkz. B\u00f6l\u00fcm 3.6 ve Cosmelli ve di\u011ferleri, 2007; Froese, 2017; Kiverstein, 2020; Thompson, 2007; Varela ve di\u011ferleri, 1993).<\/p>\n\n\n\n

Bu \u00f6zelliklerin bilin\u00e7li YZ olas\u0131l\u0131\u011f\u0131yla ili\u015fkisi, senaryodan senaryoya de\u011fi\u015fir. \u0130\u015flemleme, bilin\u00e7 i\u00e7in yeterliyse sanal bedenler ve sanal \u00e7evreler ya da biyolojik olmayan robotik bedenler de yeterli olmal\u0131d\u0131r.<\/strong> Substrata ba\u011f\u0131ml\u0131 senaryolarda bedenlenmi\u015flik ve g\u00f6m\u00fcl\u00fcl\u00fck, ilgili i\u015flevsel \u00f6rg\u00fctlenme \u00f6r\u00fcnt\u00fclerini m\u00fcmk\u00fcn k\u0131lmak i\u00e7in gerekli olabilir. \u00d6rne\u011fin fiziksel olarak bedenlenmi\u015f robotlar\u0131n, ger\u00e7ek ama biyolojik olmayan t\u00fcrden homeostaz ve allostaz\u0131 (otopoiesis olmaks\u0131z\u0131n) g\u00fcndeme getiren enerjiyle ilgili [energetic<\/em>] ve yap\u0131sal b\u00fct\u00fcnl\u00fck gereksinimleri olabilir.<\/p>\n\n\n\n

Genel olarak substrat\u0131n \u00f6zelliklerine ba\u011f\u0131ml\u0131l\u0131k artt\u0131k\u00e7a bedenlenmi\u015flik ve g\u00f6m\u00fcl\u00fcl\u00fc\u011f\u00fcn, do\u011fru t\u00fcrden dinamikleri ya da i\u015flemlemesel olmayan i\u015flevleri m\u00fcmk\u00fcn k\u0131lacak zorunlu ko\u015fullar olabilece\u011fi fikri de daha akla yatk\u0131n g\u00f6r\u00fcn\u00fcr. Biyolojik do\u011falc\u0131l\u0131k lehine ileri s\u00fcrd\u00fc\u011f\u00fcm arg\u00fcmanlar yerindeyse bedenlenmi\u015flik ve g\u00f6m\u00fcl\u00fcl\u00fck muhtemelen \u00e7ok \u00f6nemlidir ve hatta zorunlu bile olabilir.<\/strong> Bunun nedeni, bu \u00f6zelliklerin, belirtti\u011fim arg\u00fcmanlar\u0131 da g\u00fcd\u00fcleyen hayatta kalma buyru\u011funun merkezinde yer almas\u0131d\u0131r (ayr\u0131ca bkz. Aru ve di\u011ferleri, 2023). Ancak bilincin biyolojik bir \u00f6zellik olarak d\u00fc\u015f\u00fcn\u00fclmesine ra\u011fmen bedenlenmi\u015flik ile g\u00f6m\u00fcl\u00fcl\u00fc\u011f\u00fcn gerekli olmamas\u0131 da m\u00fcmk\u00fcnd\u00fcr. Beyin organoidleri, bedenlenmi\u015flik ve g\u00f6m\u00fcl\u00fcl\u00fckten tipik olarak yoksun olmalar\u0131na ra\u011fmen, bu bak\u0131mdan ilgin\u00e7 bir s\u0131nama durumu sunar (Bayne ve di\u011ferleri, 2020; Jeziorski ve di\u011ferleri, 2023).[40]<\/a><\/p>\n\n\n\n

Bu ba\u011flamdaki ba\u015fkaca sorular da derinlemesine incelenmeye de\u011fer. Bunlardan biri eyleyicilikle<\/em> [agency<\/em>] ilgilidir. Biyolojik sistemlerde eyleyicilik, (eyleyici YZ\u2019de oldu\u011fu gibi) ama\u00e7lara hizmet eden eylemleri ger\u00e7ekle\u015ftirme yetene\u011finden \u00e7o\u011fu zaman daha fazlas\u0131 anlam\u0131na gelir. Biyolojik sistemlerde ilgili hedefler tipik olarak d\u0131\u015far\u0131dan dayat\u0131lm\u0131\u015f [exogenously imposed<\/em>] de\u011fil i\u00e7 kaynakl\u0131d\u0131r [endogenous<\/em>]<\/strong> (Jaeger ve di\u011ferleri, 2024).<\/strong> Bu anlamda bir sistemin eyleyicili\u011fe sahip olmas\u0131 ya da bunu sergilemesi i\u00e7in zengin bir bedenlenmi\u015flik ve g\u00f6m\u00fcl\u00fcl\u00fck gerekli olabilir (Mitchell, 2023) ve\/veya eyleyicilik biyolojik maddenin i\u00e7inde derinden k\u00f6k salm\u0131\u015f olabilir (Levin, 2023). T\u0131pk\u0131 zek\u00e2da oldu\u011fu gibi, YZ \u201ceyleyicileri\u201d ve fiziksel robotlar, biyolojik sistemlerde bulunan g\u00fc\u00e7l\u00fc ve i\u00e7 kaynakl\u0131 t\u00fcrlere kar\u015f\u0131l\u0131k, zay\u0131f ve g\u00f6zlemciye ba\u011f\u0131ml\u0131 eyleyicilik t\u00fcrleriyle s\u0131n\u0131rlanm\u0131\u015f olabilir (bkz. B\u00f6l\u00fcm 3.8).<\/p>\n\n\n\n

<\/div>\n\n\n\n

5.8 \u00d6zet<\/em><\/p>\n\n\n\n

<\/div>\n\n\n\n

Yukar\u0131da g\u00f6zden ge\u00e7irilen senaryolar i\u00e7inden biyolojik do\u011falc\u0131l\u0131k lehine bir sav ortaya koymaya \u00e7al\u0131\u015ft\u0131m. \u00d6zellikle de ya\u015fayan maddenin, baz\u0131 \u00fcst d\u00fczey i\u015flevsel \u00f6rg\u00fctlenmeler i\u00e7in yaln\u0131zca zemin sa\u011flayan bir ko\u015ful olman\u0131n \u00f6tesinde, substrat ile i\u015flev, madde ile zihin ve ya\u015fam ile bilin\u00e7 aras\u0131nda derin bir s\u00fcreklili\u011fi vurgulayan daha g\u00fc\u00e7l\u00fc bir tonunu savundum. Bu g\u00f6r\u00fc\u015fe g\u00f6re bilin\u00e7li YZ\u2019nin, karbon temelli olmak zorunda olmasa bile, (ilgili bir anlamda) \u201cya\u015fayan\u201d YZ olmas\u0131 gerekir.<\/strong> Burada herhangi bir neo-vitalizm tortusu s\u00f6z konusu de\u011fildir. Nas\u0131l f\u0131rt\u0131na olmadan r\u00fczg\u00e2r olabiliyorsa, bilin\u00e7 olmadan da ya\u015fam olabilir.<\/p>\n\n\n\n

Bununla birlikte bu senaryolar\u0131n g\u00f6reli ikna ediciliklerini g\u00fcvenilir bi\u00e7imde de\u011ferlendirmek zordur. Ortaya koydu\u011fum arg\u00fcman kusursuz de\u011fildir. Ayr\u0131nt\u0131larda yanl\u0131\u015f olabilir, hatta t\u00fcmden yanl\u0131\u015f olabilir. Burada \u00e7e\u015fitli senaryolar\u0131 serimlemenin iki temel amac\u0131 vard\u0131r. Birincisi, bilin\u00e7li YZ\u2019nin ka\u00e7\u0131n\u0131lmaz oldu\u011funu varsaymamak gerekti\u011fini hat\u0131rlat\u0131r. \u0130kincisi, biyolojik bilin\u00e7ten sorumlu d\u00fczenekleri daha iyi anlaman\u0131n aciliyetini vurgular. Biyolojik do\u011falc\u0131l\u0131k, yeterli ko\u015fullar bak\u0131m\u0131ndan yanl\u0131\u015f \u00e7\u0131ksa bile, bilin\u00e7 olgusunun insanlar ve di\u011fer hayvanlarda ne oldu\u011funun anla\u015f\u0131lmas\u0131 i\u00e7in biyolojik bir bak\u0131\u015f a\u00e7\u0131s\u0131 yine de gerekli olabilir.<\/strong> Ve var oldu\u011fundan makul \u00f6l\u00e7\u00fcde emin olabildi\u011fimiz yerlerde bilin\u00e7 hakk\u0131nda ne kadar \u00e7ok \u015fey bilirsek, ba\u015fka yerlerde de aya\u011f\u0131m\u0131z\u0131 o kadar sa\u011flam bir zemine basar\u0131z.<\/strong><\/p>\n\n\n\n

<\/div>\n\n\n\n

6.0 Ne yap(ma)amal\u0131y\u0131z?<\/strong><\/p>\n\n\n\n

<\/div>\n\n\n\n

Bilin\u00e7li YZ yaratmak i\u00e7in ne gerekti\u011fini bilmiyoruz. Bunun ger\u00e7ekle\u015fmesini kesin olarak nas\u0131l \u00f6nleyece\u011fimizi de bilmiyoruz. Bilin\u00e7li YZ\u2019nin substrat \u00f6zelliklerini ya\u015fayan sistemlerle payla\u015fmas\u0131 gerekece\u011fi y\u00f6n\u00fcnde bir sav ortaya koydum, ama yan\u0131l\u0131yor olabilirim. Di\u011fer kuramlara dayanan ba\u015fka \u00f6neriler de yanl\u0131\u015f olabilir. \u0130nsanlarda ve (bir ihtimal) baz\u0131 di\u011fer hayvanlarda \u00fczerinde uzla\u015f\u0131lan \u00f6rnekler d\u0131\u015f\u0131nda bilincin varl\u0131\u011f\u0131n\u0131 g\u00fcvenilir bi\u00e7imde s\u0131nayacak testlerden de yoksunuz (Bayne ve di\u011ferleri, 2024; Birch, 2024).[41]<\/a> Bu ko\u015fullar alt\u0131nda etik de\u011ferlendirme yapmak zordur. Ancak YZ\u2019nin h\u0131zla geli\u015fip yayg\u0131nla\u015ft\u0131\u011f\u0131 d\u00fc\u015f\u00fcn\u00fcld\u00fc\u011f\u00fcnde bu de\u011ferlendirmelere ihtiya\u00e7 giderek artmaktad\u0131r<\/strong> (Long ve di\u011ferleri, 2024; Schwitzgebel & Garza, 2023).<\/strong><\/p>\n\n\n\n

Bu de\u011ferlendirmeleri \u00e7er\u00e7evelerken, bilin\u00e7li olan YZ sistemleri ile yaln\u0131zca bilin\u00e7li gibi g\u00f6r\u00fcnen sistemleri birbirinden ay\u0131rmak elzemdir. \u0130kincisi birincisinden daha olas\u0131d\u0131r ve her iki durumda birbirinden farkl\u0131 etik sorunlar ortaya \u00e7\u0131kar.<\/strong><\/p>\n\n\n\n

<\/div>\n\n\n\n

6.1 Ger\u00e7ek yapay bilin\u00e7<\/em><\/p>\n\n\n\n

<\/div>\n\n\n\n

Ger\u00e7ek yapay bilin\u00e7 \u00e7ok d\u00fc\u015f\u00fck bir olas\u0131l\u0131k ya da b\u00fcsb\u00fct\u00fcn imk\u00e2ns\u0131z olabilir. Ancak ister tasar\u0131m yoluyla ister kazara bu t\u00fcr sistemlerin ortaya \u00e7\u0131kmas\u0131 durumunda, bizi etik bir felaket bekleyecektir. Bilin\u00e7li olman\u0131n ahlaki stat\u00fc kazand\u0131rd\u0131\u011f\u0131n\u0131 kabul eden uzun ge\u00e7mi\u015fe sahip bir felsefi gelenek vard\u0131r (Singer, 1975). Bilin\u00e7le birlikte, olumsuz de\u011ferlili\u011fe sahip bilin\u00e7li deneyimlerin, yani ac\u0131n\u0131n, potansiyeli de ortaya \u00e7\u0131kar.<\/strong> Ger\u00e7ek yapay bilin\u00e7 yaratmak, yeni ac\u0131 t\u00fcrlerinin kitlesel \u00f6l\u00e7ekte d\u00fcnyaya getirilmesi riskini ta\u015f\u0131r.<\/strong> Sorun, insanmerkezci \u00f6nyarg\u0131lar\u0131m\u0131zla yeterince uyumlu olmayan bi\u00e7imlerde d\u0131\u015favuruldu\u011fu takdirde, bu ac\u0131n\u0131n kendisini ac\u0131 olarak bile tan\u0131yamayabilece\u011fimiz ger\u00e7e\u011fiyle daha da karma\u015f\u0131kla\u015f\u0131r.<\/strong> Buna kar\u015f\u0131n, olumsuz de\u011ferlili\u011fe sahip deneyimi \u00f6nleyecek bi\u00e7imde ger\u00e7ek yapay bilin\u00e7ler tasarlayabilece\u011fimiz ileri s\u00fcr\u00fclebilir. Fakat bunu g\u00fcvence alt\u0131na almak ya \u00e7ok g\u00fc\u00e7 ya da b\u00fcsb\u00fct\u00fcn imk\u00e2ns\u0131z olacakt\u0131r. \u00dcstelik bilincin t\u00fcm bi\u00e7imlerinin, de\u011ferlilik i\u00e7ersin i\u00e7ermesin, her h\u00e2l\u00fck\u00e2rda ahlaki stat\u00fc kazand\u0131rd\u0131\u011f\u0131 ileri s\u00fcr\u00fclebilir.<\/strong>[42]<\/a><\/p>\n\n\n\n

Ba\u015fka kayg\u0131lar da vard\u0131r. Ger\u00e7ek yapay bilin\u00e7, tasar\u0131mc\u0131lar\u0131n\u0131n y\u00fcklediklerinin \u00f6tesine ge\u00e7erek, sisteme kendi \u00e7\u0131karlar\u0131n\u0131 ve de\u011ferlerini kazand\u0131rabilir. Bu durum, YZ\u2019nin insan \u00e7\u0131karlar\u0131 ve de\u011ferleriyle uyumlu davranmas\u0131n\u0131 g\u00fcvenceye almakla ilgili zaten \u00e7etrefil \u201cde\u011fer hizalama problemini\u201d [value alignment problem<\/em>] daha da \u00e7etrefil k\u0131lar (Russell & Norvig, 2003). \u0130nsan d\u00fczeyindeki yetenekleri baz\u0131 ya da t\u00fcm alanlarda a\u015fabilecek YZ sistemleri i\u00e7in bile, potansiyel olarak ayr\u0131\u015fan \u00e7\u0131karlar\u0131n uyumlu kalmas\u0131n\u0131 sa\u011flamak gibi pratik bir g\u00fc\u00e7l\u00fck vard\u0131r. Ayr\u0131ca YZ\u2019nin kendi \u00e7\u0131karlar\u0131na sahip olmas\u0131 me\u015fru kabul edildi\u011finde, bu pratik soruna ne zaman ve nas\u0131l m\u00fcdahale etmenin do\u011fru olaca\u011f\u0131 da ahlaki bir g\u00fc\u00e7l\u00fck olarak kar\u015f\u0131m\u0131za \u00e7\u0131kar.[43]<\/a><\/p>\n\n\n\n

Ger\u00e7ek yapay bilincin salt kendisi u\u011fruna geli\u015ftirilmesi a\u00e7\u0131k bir hedef olmamal\u0131d\u0131r.<\/strong>[44]<\/a> Bu s\u00f6yledi\u011fim bariz g\u00f6r\u00fcnebilir, fakat \u015fa\u015f\u0131rt\u0131c\u0131 bi\u00e7imde \u201cbilin\u00e7li YZ\u201dye s\u0131kl\u0131kla [YZ\u2019nin] b\u00fcy\u00fck \u00fclk\u00fc[s\u00fc] (kutsal kasesi) [holy grail<\/em>] muamelesi yap\u0131lmaktad\u0131r.<\/strong> Oysa teknolojiler geli\u015ftik\u00e7e, biyolojik bilin\u00e7 anlay\u0131\u015f\u0131m\u0131z derinle\u015ftik\u00e7e ve testlerimiz iyile\u015ftik\u00e7e bilincin varl\u0131\u011f\u0131na ili\u015fkin inan\u0131\u015flar\u0131m\u0131z\u0131 [credence<\/em>] s\u00fcrekli g\u00f6zden ge\u00e7irmemiz gerekir.<\/strong> De\u011fi\u015fip geli\u015fen bu inan\u0131\u015flar\u0131 da yapay zek\u00e2y\u0131 geli\u015ftirme \u00e7abalar\u0131n\u0131 bilincin ger\u00e7ekle\u015fme olas\u0131l\u0131\u011f\u0131ndan uzak tutacak \u015fekilde y\u00f6nlendirmek i\u00e7in kullanmal\u0131y\u0131z. <\/strong>Ger\u00e7ek yapay bilincin mevcut gidi\u015fatlar\u0131 i\u00e7in imk\u00e2ns\u0131z olabilece\u011fi olgusu, bu fikri yeterince d\u00fc\u015f\u00fcn\u00fclmemi\u015f bir teknolojik kopu\u015f [techno-rapture<\/em>] u\u011fruna k\u00f6r\u00fc k\u00f6r\u00fcne takip etmemizi me\u015fru k\u0131lmaz.<\/p>\n\n\n\n

Bu noktada dile getirilen itirazlarla, toplumsal bak\u0131mdan yararl\u0131 bir YZ geli\u015ftirmenin, bilin\u00e7le ili\u015fkilendirilen kimi i\u015flevlerin uygulanmas\u0131yla g\u00fc\u00e7lendirilebilece\u011fi g\u00f6r\u00fc\u015f\u00fc aras\u0131nda bir gerilim ortaya \u00e7\u0131kar. Ger\u00e7ekten de b\u00f6yle d\u00fc\u015f\u00fcnmek i\u00e7in gerek\u00e7elerimiz var. Mevcut YZ\u2019nin ba\u015far\u0131s\u0131zl\u0131k sahalar\u0131n\u0131n bir b\u00f6l\u00fcm\u00fcn\u00fcn, bilince yak\u0131ndan ba\u011fl\u0131 g\u00f6r\u00fcnen insan kabiliyetleriyle kesi\u015fti\u011fi \u00f6ne s\u00fcr\u00fclm\u00fc\u015ft\u00fcr. \u00d6rnekler aras\u0131nda yeni durumlara h\u0131zla genelleme yapma, etkili \u00fcst-bili\u015f [metacognition<\/em>] ve az miktarda veriden \u00f6\u011frenme say\u0131labilir (Fleming, 2021; VanRullen & Kanai, 2021; Vong ve di\u011ferleri, 2024). B\u00fcy\u00fck resme bak\u0131ld\u0131\u011f\u0131nda, bu zorluklarla etkili bi\u00e7imde ba\u015fa \u00e7\u0131kabilen YZ sistemlerinin istenmesi s\u00f6z konusu olabilir.<\/p>\n\n\n\n

\u0130nce bir \u00e7izgi \u00fczerinde y\u00fcr\u00fcnmesi gerekiyor. Yararl\u0131 i\u015flevlerin, insandan (ya da daha genel olarak biyolojik \u00f6rneklerden) yeterince farkl\u0131 yollardan uygulanabilece\u011fi ve b\u00f6ylece fark\u0131nda olmadan bilinci \u00f6rnekleme riskinin en aza indirilebilece\u011fi d\u00fc\u015f\u00fcncesi akla yatk\u0131n g\u00f6r\u00fcnmektedir.<\/strong> \u201cYeterince farkl\u0131\u201d say\u0131lacak \u015fey, hangi bilin\u00e7 kuram\u0131 ve hangi yapay bilin\u00e7 senaryosunun tercih edildi\u011fine ba\u011fl\u0131 olarak de\u011fi\u015fecektir.<\/strong> Biyolojik bilincin do\u011fas\u0131n\u0131 daha iyi anlamak i\u00e7in i\u015flemlemesel modeller kullanan bilim insanlar\u0131n\u0131n bile benzer bir ince \u00e7izgide y\u00fcr\u00fcmesi gerekmesi dikkat \u00e7ekicidir<\/strong> (Seth, 2009).<\/strong> Burada da \u00e7izginin nereye \u00e7ekilece\u011fi kuramsal tercihlere ba\u011fl\u0131 olacakt\u0131r. Fakat ger\u00e7ek yapay bilincin salt kendisi u\u011fruna geli\u015ftirilmesi \u00e7ok k\u00f6t\u00fc bir fikir olarak kalmaya devam edecektir (Seth, 2023).<\/p>\n\n\n\n

<\/div>\n\n\n\n

6.2 Bilin\u00e7liymi\u015f gibi g\u00f6r\u00fcnen YZ<\/em><\/p>\n\n\n\n

<\/div>\n\n\n\n

Ger\u00e7ek yapay bilincin etraf\u0131n\u0131 saran inand\u0131r\u0131c\u0131l\u0131k eksikli\u011fi ve belirsizli\u011fin aksine, YZ sistemlerinin yak\u0131nda bize bilin\u00e7li olduklar\u0131na dair g\u00fc\u00e7l\u00fc izlenimler verecekleri neredeyse kesindir.<\/strong> En yeni LLM\u2019lere verilen tepkilerin g\u00f6sterdi\u011fi \u00fczere bu t\u00fcr sistemler \u015fimdiden aram\u0131zda olabilir, \u00f6zellikle de bu LLM\u2019ler bize kendi \u201cbilin\u00e7lerinden\u201d s\u00f6z ettiklerinde (Shanahan & Singler, 2024). Yeterince ikna edici LLM\u2019ler, geli\u015fmi\u015f \u00fcretken [generative<\/em>] g\u00f6r\u00fcnt\u00fc ve ses avatarlar\u0131yla (ve daha da \u00f6tesine ge\u00e7ersek robotlarla) birle\u015ftirildi\u011finde, onlar hakk\u0131nda ne bildi\u011fimiz ya da neye inand\u0131\u011f\u0131m\u0131zdan ba\u011f\u0131ms\u0131z olarak, bilin\u00e7li olduklar\u0131n\u0131 hissetmekten kendimizi alamayaca\u011f\u0131m\u0131z yapay sistemlerle d\u00fcnyay\u0131 payla\u015faca\u011f\u0131z.<\/p>\n\n\n\n

Yapay bilince ili\u015fkin izlenimlerin d\u00fczeltilemez olabilece\u011fini vurgulamak istiyorum.<\/strong> Baz\u0131 g\u00f6rsel yan\u0131lsamalar bili\u015fsel bak\u0131mdan ge\u00e7irimsizdir [impenetrable<\/em>]. M\u00fcller-Lyer yan\u0131lsamas\u0131ndaki iki \u00e7izgi, ayn\u0131 uzunlukta olduklar\u0131n\u0131 bilseniz bile, her zaman farkl\u0131 g\u00f6r\u00fcn\u00fcr (\u015eekil 1). Yeterince geli\u015fmi\u015f YZ sistemleri, bili\u015fsel bak\u0131mdan ge\u00e7irimsiz bilin\u00e7 yan\u0131lsamalar\u0131na yol a\u00e7abilir. <\/strong>Bu vurgulamaya de\u011fer bir noktad\u0131r, \u00e7\u00fcnk\u00fc aksi h\u00e2lde, uzman bilgisinin ya da sars\u0131lmaz inan\u00e7lar\u0131m\u0131z\u0131n bizi, bilin\u00e7siz sistemleri bilin\u00e7liymi\u015f gibi hissetmekten ya da bunun sonu\u00e7lar\u0131ndan ba\u011f\u0131\u015f\u0131k k\u0131laca\u011f\u0131n\u0131 iyimser bir \u015fekilde varsayabiliriz.<\/strong> Buna kar\u015f\u0131l\u0131k, ne bildi\u011fimiz veya neye inand\u0131\u011f\u0131m\u0131zdan ba\u011f\u0131ms\u0131z olarak, YZ sistemlerinin hepimiz i\u00e7in bilin\u00e7liymi\u015f gibi g\u00f6r\u00fcnmesi b\u00fcy\u00fck olas\u0131l\u0131kla ka\u00e7\u0131n\u0131lmazd\u0131r.[45]<\/a><\/p>\n\n\n\n

\"\"<\/a><\/figure>\n\n\n\n

\u015eekil 1<\/strong>. M\u00fcller-Lyer yan\u0131lsamas\u0131<\/p>\n\n\n\n

Bilin\u00e7liymi\u015f gibi g\u00f6r\u00fcnen YZ\u2019nin de kendine \u00f6zg\u00fc etik sorunlar\u0131 vard\u0131r. Bu sorunlar, ger\u00e7ek yapay bilin\u00e7le ba\u011flant\u0131l\u0131 olanlara k\u0131yasla varolu\u015fsal bak\u0131mdan daha az tehditk\u00e2r olabilir, fakat bilin\u00e7liymi\u015f gibi g\u00f6r\u00fcnen YZ\u2019nin g\u00fc\u00e7l\u00fc olas\u0131l\u0131\u011f\u0131 bunlar\u0131 yine de acil k\u0131lar.<\/p>\n\n\n\n

Bilin\u00e7li gibi g\u00f6r\u00fcnen YZ, ister kazara ister (kurumsal) tasar\u0131mla olsun, psikolojik savunmas\u0131zl\u0131klar\u0131m\u0131z\u0131 istismar etmek i\u00e7in elveri\u015fli bir konumdad\u0131r.<\/strong> Bir LLM\u2019nin bilin\u00e7li oldu\u011funu hissetti\u011fimiz i\u00e7in bizi ger\u00e7ekten anlad\u0131\u011f\u0131na ve bize ger\u00e7ekten \u00f6nem verdi\u011fine inan\u0131rsak, verdi\u011fi \u00f6neri k\u00f6t\u00fc oldu\u011funda bile onu yerine getirmeye daha meyilli olabiliriz. Bir algoritmayla de\u011fil de ba\u015fka bir zihinle muhatap oldu\u011fumuzu hissedersek, dijital mahremiyetimizden geriye kalanlardan dahi vazge\u00e7meye daha g\u00f6n\u00fcll\u00fc olabiliriz<\/strong> (V\u00e9liz, 2021).<\/strong><\/p>\n\n\n\n

Savunmas\u0131zl\u0131\u011f\u0131m\u0131za ili\u015fkin bu kayg\u0131lar ger\u00e7ek yapay bilin\u00e7 i\u00e7in de ge\u00e7erlidir. Ancak bilin\u00e7liymi\u015f gibi g\u00f6r\u00fcnen YZ\u2019ye \u00f6zg\u00fc bir husus, daha \u00f6nce film ve televizyon dizisinden hareketle \u201cWest World\u201d senaryosu diye adland\u0131rd\u0131\u011f\u0131m durumda ortaya \u00e7\u0131kar (Seth, 2021; ayr\u0131ca bkz. Chrisley, 2020; Seth, 2023). Bir sistemin bilin\u00e7li olmad\u0131\u011f\u0131na inand\u0131\u011f\u0131m\u0131z ama buna ra\u011fmen bilin\u00e7liymi\u015f gibi hissetti\u011fimiz bir durumda zor bir se\u00e7imle kar\u015f\u0131la\u015f\u0131r\u0131z. Ya sistemi \u00f6nemsemeye karar veririz, yani bilin\u00e7li olmad\u0131\u011f\u0131na inansak bile ona etik ve ahlaki stat\u00fc tan\u0131r\u0131z; ya da bilin\u00e7liymi\u015f gibi hissetmeye devam etmemize ra\u011fmen onu \u00f6nemsememeyi tercih ederiz<\/strong>.<\/p>\n\n\n\n

Se\u00e7eneklerin hi\u00e7biri iyi de\u011fildir. Bilin\u00e7siz sistemleri \u00f6nemsemeye karar verirsek, ahlaki \u00e7emberimizin s\u0131n\u0131r\u0131n\u0131 de\u011fi\u015ftirmi\u015f oluruz. \u0130lgimizi ve kaynaklar\u0131m\u0131z\u0131, bunlar\u0131 hak eden ba\u015fka \u015feylerden, ba\u015fka insanlardan ve insan-d\u0131\u015f\u0131 hayvanlardan uzakla\u015ft\u0131r\u0131r\u0131z.<\/strong> Bu sistemleri \u00f6nemsememe<\/em>ye karar verirsek, kendi zihnimizi hoyratla\u015ft\u0131rma riski do\u011far. West World\u2019de ya\u015fanan budur ve bu d\u00fc\u015f\u00fcnce hatt\u0131 en az\u0131ndan Kant\u2019\u0131n ahlak derslerine kadar uzan\u0131r. Duygular\u0131 varm\u0131\u015f gibi g\u00f6r\u00fcnen sistemlere yokmu\u015f gibi davran\u0131rsak, ba\u015fkalar\u0131n\u0131n dile getirdi\u011fi ger\u00e7ek duygulara kar\u015f\u0131 etik bak\u0131mdan duyars\u0131z h\u00e2le gelebiliriz, b\u00f6ylece ahlaki duyarl\u0131l\u0131klar\u0131m\u0131z da kabala\u015f\u0131r.<\/strong><\/p>\n\n\n\n

\u00d6lm\u00fc\u015f arkada\u015flar\u0131m\u0131z\u0131n ya da akrabalar\u0131m\u0131z\u0131n ki\u015fisel verileriyle e\u011fitilmi\u015f LLM destekli avatarlar\u0131n yer ald\u0131\u011f\u0131 \u00e7arp\u0131c\u0131 bir distopik senaryo d\u00fc\u015f\u00fcn\u00fclebilir. Bu size uzak bir ihtimal gibi g\u00f6r\u00fcnebilir, ancak bu t\u00fcr fikirler \u015fimdiden dola\u015f\u0131mdad\u0131r.[46]<\/a> Bu senaryo, beraberinde saymakla bitmeyecek sorunlar getirecektir, yas tutman\u0131n ve unutman\u0131n psikolojik \u00f6nemine dair ola\u011fan\u00fcst\u00fc bir safdillik, otobiyografik an\u0131lar\u0131n \u00e7arp\u0131t\u0131lmas\u0131, insanlar\u0131n istismar edilmesi i\u00e7in f\u0131rsatlar sunmas\u0131 bunlardan birka\u00e7\u0131d\u0131r (\u00f6len baban\u0131z\u0131n, ya da daha beteri, \u00f6len k\u0131z\u0131n\u0131z\u0131n size \u201cbu ola\u011fan\u00fcst\u00fc i\u015f f\u0131rsat\u0131na yat\u0131r\u0131m yapman\u0131 ger\u00e7ekten istiyorum\u201d dedi\u011fini bir d\u00fc\u015f\u00fcn\u00fcn).<\/p>\n\n\n\n

Burada da y\u00fcr\u00fcnmesi gereken ince bir \u00e7izgi var. Bilin\u00e7li g\u00f6r\u00fcnen YZ baz\u0131 yararlar getirebilir. Etkile\u015fim kurmay\u0131 kolayla\u015ft\u0131rarak verimlili\u011fi art\u0131rabilir ve eri\u015filebilirli\u011fi geni\u015fletebilir. Bireyselle\u015ftirilmi\u015f terapi ve ilk bak\u0131\u015fta pek cazip gelmese de yaln\u0131zl\u0131\u011f\u0131n hafifletilmesi gibi alanlarda f\u0131rsatlar sunabilir (Stade ve di\u011ferleri, 2024). \u0130lgi \u00e7ekici bir se\u00e7enek, LLM\u2019leri bir t\u00fcr \u201crol yapma\u201d kapsam\u0131nda ele almakt\u0131r (Shanahan ve di\u011ferleri, 2023). Bu strateji yararl\u0131 bir psikolojik mesafe koysa bile bizi ancak bir yere kadar g\u00f6t\u00fcrecektir.<\/p>\n\n\n\n

Bilin\u00e7liymi\u015f gibi g\u00f6r\u00fcnen YZ\u2019nin geli\u015fini ka\u00e7\u0131n\u0131lmaz olarak tarif ettim. Fakat bu kolektif bir \u015fekilde verece\u011fimiz kararlara ba\u011fl\u0131d\u0131r. <\/strong>Alternatif gelecekler de m\u00fcmk\u00fcnd\u00fcr. Bilin\u00e7li g\u00f6r\u00fcnen YZ\u2019ler iyi gerek\u00e7elendirilmi\u015f kullan\u0131m durumlar\u0131yla s\u0131n\u0131rlanabilir ve YZ\u2019nin b\u00fcy\u00fck k\u0131sm\u0131, psikolojik \u00f6nyarg\u0131lar\u0131m\u0131zdan yararlanmak yerine onlara diren\u00e7 g\u00f6sterecek \u015fekilde tasarlanabilir. Daniel Dennett\u2019in dedi\u011fi gibi, \u201cKendi \u00f6mr\u00fcm\u00fczde g\u00f6r\u00fcp g\u00f6rebilece\u011fimiz yeg\u00e2ne \u015feyler zeki ara\u00e7lard\u0131r, meslekta\u015flar de\u011fil. Onlar\u0131 meslekta\u015flar\u0131n\u0131z olarak d\u00fc\u015f\u00fcnmeyin, meslekta\u015fa d\u00f6n\u00fc\u015ft\u00fcrmeye \u00e7al\u0131\u015fmay\u0131n ve en \u00f6nemlisi, onlar\u0131n meslekta\u015f\u0131n\u0131z oldu\u011funa dair kendinizi kand\u0131rmak gibi bir hataya d\u00fc\u015fmeyin.\u201d<\/strong>[47]<\/a><\/p>\n\n\n\n

<\/div>\n\n\n\n

7.0 Sonu\u00e7lar<\/strong><\/p>\n\n\n\n

<\/div>\n\n\n\n

Biz insanlar, kendi suretimizde teknolojiler yaratma ve kendimizi yaratt\u0131\u011f\u0131m\u0131z bu teknolojilere yans\u0131tma al\u0131\u015fkanl\u0131\u011f\u0131na sahibiz.<\/strong> Bilin\u00e7 ile YZ aras\u0131ndaki ili\u015fkiyi ku\u015fatan kavramsal sis bulutu, bu teknolojinin temsil etti\u011fi tehditleri ve f\u0131rsatlar\u0131 berrak bir \u015fekilde g\u00f6rmemizi engeller.<\/strong> Ayr\u0131ca kendimizi<\/em> de a\u00e7\u0131k se\u00e7ik g\u00f6rmemizi engeller. Biz ya\u015fayan, nefes alan, hisseden, eyleyen, d\u00fc\u015f\u00fcnen ve deneyimleyen varl\u0131klar\u0131z. Biyolojik beyinlerin ve insan deneyiminin zenginli\u011fini, \u201cdeepfake\u201d destekli dil modellerinin enformasyon-i\u015fleme d\u00fczenekleriyle bir ve ayn\u0131 \u015fey sayarsak, makineleri abart\u0131r, kendimizi k\u00fc\u00e7\u00fcmseriz.<\/strong>[48]<\/a><\/p>\n\n\n\n

\u0130nsanmerkezci ve insanbi\u00e7imci \u00f6nyarg\u0131lar\u0131m\u0131z\u0131n g\u00fcc\u00fcne ra\u011fmen, YZ ak\u0131lland\u0131k\u00e7a bilin\u00e7 kendili\u011finden e\u015flik etmeyecektir.<\/strong> Bilincin geleneksel YZ sistemlerinde hi\u00e7 m\u00fcmk\u00fcn olmayabilece\u011fini d\u00fc\u015f\u00fcnmek i\u00e7in g\u00fc\u00e7l\u00fc nedenlerimiz vard\u0131r. Dahas\u0131, bilincin ya\u015fayan sistemlerin bir \u00f6zelli\u011fi olabilece\u011fini ve hatta belki yaln\u0131zca onlara \u00f6zg\u00fc olabilece\u011fini varsaymak i\u00e7in de (neo-vitalizme d\u00fc\u015fmeyen) baz\u0131 g\u00fc\u00e7l\u00fc nedenler mevcuttur.<\/strong> Bildi\u011fimiz b\u00fct\u00fcn bilin\u00e7li sistem \u00f6rneklerinin biyolojik ve canl\u0131 oldu\u011funu kaydetmek bile, biyolojik do\u011falc\u0131l\u0131\u011f\u0131n i\u015flemlemeselci i\u015flevselcilikten daha spek\u00fclatif olmad\u0131\u011f\u0131n\u0131 g\u00f6sterir.<\/p>\n\n\n\n

\u00d6zellikle biyolojik do\u011falc\u0131l\u0131k lehindeki arg\u00fcmanlarda pek \u00e7ok belirsizlik s\u00fcrmektedir. Burada bilincin ya\u015famla ili\u015fkisini ve daha \u00f6zel olarak \u00f6ng\u00f6r\u00fcc\u00fc i\u015fleme, serbest enerji ilkesi, otopoiesis ve metabolizman\u0131n termodinami\u011fi aras\u0131ndaki ba\u011flant\u0131lar\u0131 ayd\u0131nlatmak i\u00e7in heyecan verici imk\u00e2nlar bulunmaktad\u0131r. Sinirbi\u00e7imsel ve sentetik biyoloji teknolojilerinin, ger\u00e7ek yapay bilin\u00e7 imk\u00e2n\u0131na dair ibreyi ne y\u00f6nde ve nas\u0131l etkiledi\u011fini anlamak i\u00e7in de \u00f6n\u00fcm\u00fczde \u00e7ok i\u015f vard\u0131r. Bu teknolojiler donan\u0131m ile yaz\u0131l\u0131m aras\u0131ndaki ayr\u0131m\u0131 gitgide eritmekte, bununla birlikte zihin yaz\u0131l\u0131m\u0131 [mindware<\/em>] ile \u0131slak donan\u0131m [wetware<\/em>] aras\u0131ndaki ili\u015fkiye yakla\u015fmaktad\u0131r.<\/strong> Ayr\u0131ca Turing makinesinin kanonik \u00e7er\u00e7evesinin \u00f6tesindeki daha geni\u015f i\u015flemleme bi\u00e7imlerini dikkate alman\u0131n \u00f6nemini de vurgularlar.[49]<\/a><\/p>\n\n\n\n

Bu belirsizliklere ra\u011fmen bilin\u00e7li bir etik tutum benimsemenin \u00f6nemi, bir ba\u015fka be\u015fer\u00ee al\u0131\u015fkanl\u0131\u011fa dikkat \u00e7eker. Ahlaki stat\u00fcy\u00fc bilin\u00e7li \u015feylerden (pek \u00e7ok insan-d\u0131\u015f\u0131 hayvandan ve hatta bazen di\u011fer insanlardan) esirgeme y\u00f6n\u00fcndeki talihsiz sicilimiz… YZ\u2019ye bilin\u00e7 atfetmekten ka\u00e7\u0131nmak bizi yine tarihin yanl\u0131\u015f saf\u0131na m\u0131 s\u00fcr\u00fckleyecektir? <\/strong>Buradaki durumun farkl\u0131 oldu\u011funa inan\u0131yorum. Psikolojik \u00f6nyarg\u0131lar\u0131m\u0131z bizi sahte negatiflerden \u00e7ok sahte pozitiflere s\u00fcr\u00fcklemeye daha yatk\u0131nd\u0131r. \u0130nsan-d\u0131\u015f\u0131 hayvanlarla k\u0131yasland\u0131\u011f\u0131nda, b\u00fcy\u00fck dil modelleri bilin\u00e7 bak\u0131m\u0131ndan \u00f6nemli olmayan a\u00e7\u0131lardan bize daha \u00e7ok benzerlik g\u00f6sterebilir ve bilincin as\u0131l \u00f6nem ta\u015f\u0131d\u0131\u011f\u0131 a\u00e7\u0131lardan daha az benzer olabilir.<\/strong><\/p>\n\n\n\n

“Psikolojik \u00f6nyarg\u0131lar\u0131m\u0131z bizi sahte negatiflerden \u00e7ok sahte pozitiflere s\u00fcr\u00fcklemeye daha yatk\u0131nd\u0131r.”<\/p><\/blockquote><\/figure>\n\n\n\n

Dil modelleri ve \u00fcretici deepfake<\/em>\u2019ler bir s\u00fcre konu\u015fulacak, fakat yapay zek\u00e2n\u0131n y\u00fcr\u00fcy\u00fc\u015f\u00fc daha yeni ba\u015fl\u0131yor. Ger\u00e7ek yapay bilinci yaratmaktan ka\u00e7\u0131nman\u0131n kesin yollar\u0131n\u0131 bilmiyoruz ve bilin\u00e7li gibi g\u00f6r\u00fcnen sistemler b\u00fcy\u00fck olas\u0131l\u0131kla hayat\u0131m\u0131zda olacaklar.<\/strong> 1800\u2019lerin ba\u015f\u0131nda, 19 ya\u015f\u0131ndaki Mary Shelley ba\u015fyap\u0131t\u0131 Frankenstein<\/em>\u2019\u0131 yazd\u0131\u011f\u0131nda, \u00e7a\u011flar boyunca insanlar\u0131n kula\u011f\u0131na k\u00fcpe olacak bir ahlaki hik\u00e2ye [cautionary moral tale<\/em>] anlat\u0131yordu. Kitap, \u00e7o\u011fu zaman ya\u015fam yaratma k\u00fcstahl\u0131\u011f\u0131na kar\u015f\u0131 bir uyar\u0131 olarak okunmaktad\u0131r. Oysa Frankenstein\u2019\u0131n as\u0131l Prometheus\u00e7u g\u00fcnah\u0131, yarat\u0131\u011f\u0131na ya\u015fam vermesi de\u011fil, bilin\u00e7 vermesiydi. Deh\u015fet ve y\u0131k\u0131m getiren; yarat\u0131\u011f\u0131n\u0131n dayanma, ac\u0131 \u00e7ekme, k\u0131skan\u00e7l\u0131k, keder ve \u00f6fke duyma kapasitesiydi.<\/strong>[50]<\/a> Bizler, ister ger\u00e7ekte ister g\u00f6r\u00fcn\u00fc\u015fte olsun, ayn\u0131 hataya d\u00fc\u015fmemeliyiz.<\/strong><\/p>\n\n\n\n

<\/div>\n\n\n\n

Te\u015fekk\u00fcr<\/strong><\/p>\n\n\n\n

<\/div>\n\n\n\n

Jaan Aru, Adam Barrett, Tim Bayne, Ned Block, Simon Bowes, Chris Buckley, David Chalmers ve NYU haftal\u0131k felsefe seminer grubu, Shamil Chandaria, Robert Chis-Ciure, Stephen Fleming, Karl Friston, Jakob Hohwy ve Monash Centre for Consciousness and Contemplative Studies, Johannes Kleiner, Christof Koch, Nicolas Kuske, Nick Lane, Shane Legg, Alexander Lerchner, Michael Levin, Liad Mudrik, Angus Nisbet, Thomas Parr, Gualtiero Piccinini, Claudia Passos, Michael Pollan, Aza Raskin, Luke Roelofs, Adam Rostowski, Michael Silberstein, Murray Shanahan, Ray Tallis ve Sussex Centre for Consciousness Science\u2019taki meslekta\u015flar\u0131m da d\u00e2hil olmak \u00fczere pek \u00e7ok ki\u015finin c\u00f6mert yorum ve \u00f6nerileri i\u00e7in minnettar\u0131m. Yararl\u0131 yorumlar\u0131 i\u00e7in ayr\u0131ca d\u00f6rt anonim hakeme de te\u015fekk\u00fcr ederim.<\/p>\n\n\n\n

<\/div>\n\n\n\n

Destek<\/strong><\/p>\n\n\n\n

<\/div>\n\n\n\n

Bu proje, Avrupa Birli\u011fi\u2019nin Horizon 2020 ara\u015ft\u0131rma ve yenilik program\u0131 kapsam\u0131nda Avrupa Ara\u015ft\u0131rma Konseyi\u2019nden (ERC) CONSCIOUS projesi i\u00e7in 101019254 numaral\u0131 hibe s\u00f6zle\u015fmesiyle fon alm\u0131\u015ft\u0131r.<\/p>\n\n\n\n

<\/div>\n\n\n\n

\u00c7\u0131kar \u00e7at\u0131\u015fmas\u0131<\/strong><\/p>\n\n\n\n

<\/div>\n\n\n\n

Yazar Conscium Ltd ve AllJoined Inc. i\u00e7in dan\u0131\u015fmanl\u0131k hizmeti vermektedir.<\/p>\n\n\n\n

<\/div>\n\n\n\n

Dipnotlar<\/strong><\/p>\n\n\n\n

<\/div>\n\n\n\n

[1]<\/a> https:\/\/spectrum.ieee.org\/ai-agents<\/p>\n\n\n\n

[2]<\/a> \u00d6rne\u011fin, yaln\u0131zca hayal k\u0131r\u0131kl\u0131\u011f\u0131 ya\u015famaktan farkl\u0131 olarak pi\u015fmanl\u0131k deneyimleme yetisi, muhtemelen eylemlerin gelecekteki sonu\u00e7lar\u0131 \u00fczerine ak\u0131l y\u00fcr\u00fctebilmeyi ve birbiriyle rekabet eden eylem planlar\u0131 aras\u0131nda karar vermeyi gerektirir. Genel olarak zek\u00e2 ile bilincin birbirinden ba\u011f\u0131ms\u0131z g\u00f6r\u00fcn\u00fcp g\u00f6r\u00fcnmeyece\u011fi, sorunun mant\u0131ksal m\u0131, nomolojik mi (d\u00fcnyam\u0131zdaki do\u011fa yasalar\u0131na g\u00f6re, t\u00fcm m\u00fcmk\u00fcn d\u00fcnyalarda de\u011fil), yoksa ger\u00e7ekten var olan sistemler a\u00e7\u0131s\u0131ndan m\u0131 (biyolojik ya da ba\u015fka t\u00fcr sistemler) soruldu\u011funa ba\u011fl\u0131d\u0131r. Makul olan, bilin\u00e7 ile zek\u00e2n\u0131n mant\u0131ksal ve nomolojik bak\u0131mdan ba\u011f\u0131ms\u0131z oldu\u011fudur.<\/p>\n\n\n\n

[3]<\/a> Bir\u00e7ok \u00fclkede otomobillerin yan aynalar\u0131n\u0131n \u00fczerine \u015fu uyar\u0131 kaz\u0131nm\u0131\u015ft\u0131r: \u201cAynadaki cisimler g\u00f6r\u00fcnd\u00fcklerinden daha yak\u0131nd\u0131r.\u201d Bunu her zaman felsefi olarak ilgin\u00e7 bulmu\u015fumdur.<\/p>\n\n\n\n

[4]<\/a> Daha do\u011frusu Kleene (1956), McCulloch\u2013Pitts a\u011flar\u0131n\u0131n, Turing makinelerinden daha s\u0131n\u0131rl\u0131 bir i\u015flemleme g\u00fcc\u00fcne sahip olan sonlu durum otomatlar\u0131na denk oldu\u011funu g\u00f6sterdi. S\u0131n\u0131rs\u0131z bir depolama (ve okuma\u2013yazma eri\u015fimi) verildi\u011finde bunlar Turing makinelerine denk h\u00e2le gelirler. Bu noktay\u0131 vurgulayan hakeme te\u015fekk\u00fcr ederim.<\/p>\n\n\n\n

[5]<\/a> Turing, bilin\u00e7le ilgili sorular\u0131n (makine) zek\u00e2s\u0131yla ilgili sorulardan ayr\u0131 oldu\u011funu fark etmi\u015fti: \u201cBilince dair hi\u00e7bir gizem yokmu\u015f gibi bir izlenim vermek istemem\u2026 Ama bu gizemlerin, bu makalede ele ald\u0131\u011f\u0131m\u0131z soruya cevap vermeden \u00f6nce mutlaka \u00e7\u00f6z\u00fclmesi gerekti\u011fini de d\u00fc\u015f\u00fcnm\u00fcyorum\u201d (Turing, 1950, s. 446). Bu k\u0131ymetli al\u0131nt\u0131 i\u00e7in Jakob Hohwy\u2019ye te\u015fekk\u00fcr ederim.<\/p>\n\n\n\n

[6]<\/a> Turing i\u015flemlemesinin kapsam\u0131na ili\u015fkin ba\u015fka bir bak\u0131\u015f a\u00e7\u0131s\u0131 \u201cfiziksel Church\u2013Turing tezi\u201d (PCTT) taraf\u0131ndan sa\u011flan\u0131r. PCTT\u2019nin birden \u00e7ok yorumu vard\u0131r ve hepsi h\u00e2len birer varsay\u0131md\u0131r. Bu makaledeki arg\u00fcmanlar\u0131m, \u201c\u0131l\u0131ml\u0131\u201d bir PCTT ile uyumludur. Bu g\u00f6r\u00fc\u015fe g\u00f6re fiziksel bir sistem taraf\u0131ndan i\u015flemlenebilir olan her fonksiyon, bir Turing makinesi taraf\u0131ndan da i\u015flemlenebilirdir. Bu g\u00f6r\u00fc\u015f, \u201ccesur\u201d PCTT yorumlar\u0131ndan ayr\u0131d\u0131r. O yorumlar (bana g\u00f6re ikna edici olmayan bi\u00e7imde) fiziksel bir sistemin yapt\u0131\u011f\u0131 her \u015feyin bir Turing makinesi taraf\u0131ndan i\u015flemlenebilir oldu\u011funu ileri s\u00fcrer. Bkz. N. G. Anderson & G. Piccinini, 2024; Chalmers, 1996b.<\/p>\n\n\n\n

[7]<\/a> Shannon \u00e7er\u00e7evesinde enformasyonu, belirsizli\u011fi niceliksel olarak betimleme y\u00f6ntemi \u015feklinde d\u00fc\u015f\u00fcnmek en uygunudur (Shannon & Weaver, 1949). Enformasyon kuram\u0131, uyguland\u0131\u011f\u0131 her \u015feyin \u201cenformasyon i\u015fledi\u011fini\u201d varsaymadan da geni\u015f \u00f6l\u00e7ekte uygulanabilir, t\u0131pk\u0131 baz\u0131 \u015feyler i\u015flemlemeli olarak modellenebildi\u011fi h\u00e2lde kendilerinin i\u015flemlemesel olmamas\u0131 gibi. Bkz. N. M. Anderson & G. Piccinini, 2024; Chalmers, 1994; Floridi, 2010.<\/p>\n\n\n\n

[8]<\/a> https:\/\/www.vox.com\/future-perfect\/351893\/consciousness-ai-machines-neuroscience-mind<\/a><\/p>\n\n\n\n

[9]<\/a> \u00c7oklu ger\u00e7ekle\u015febilirli\u011fin tam olarak ne anlama geldi\u011fi (Polger & Shapiro, 2016), bunun nas\u0131l s\u0131nanabilece\u011fi (Shapiro, 2008) ve substrat ba\u011f\u0131ms\u0131zl\u0131\u011f\u0131 (ve esnekli\u011fi) ile ili\u015fkisi \u00fczerine canl\u0131 bir felsefi tart\u0131\u015fma y\u00fcr\u00fcmektedir. Polger ve Shapiro\u2019ya g\u00f6re \u00e7oklu ger\u00e7ekle\u015febilirlik, \u201cilgili bak\u0131mlardan ayn\u0131 i\u015flevin, ilgili bak\u0131mlardan farkl\u0131 yollarla yerine getirilmesi\u201d durumunda s\u00f6z konusudur (2016, s. 77). Sezgisel olarak substrat esnekli\u011fi, yaln\u0131zca ayn\u0131 substratlar i\u00e7inde de\u011fil substratlar aras\u0131nda da i\u015flevsel \u00f6rg\u00fctlenmenin de\u011fi\u015fmezli\u011fini gerektiren \u201cdaha g\u00fc\u00e7l\u00fc\u201d bir \u00e7oklu ger\u00e7ekle\u015febilirlik versiyonu gibi d\u00fc\u015f\u00fcn\u00fclebilir. Ancak Polger ve Shapiro\u2019nun anlay\u0131\u015f\u0131nda, alternatif substratlar t\u00fcm ilgili a\u00e7\u0131lardan nedensel olarak izomorfik ise (ve \u00f6yle olmak zorundaysa), hi\u00e7bir \u00e7oklu ger\u00e7ekle\u015febilirlik olmaks\u0131z\u0131n da substrat esnekli\u011fi mant\u0131ksal olarak m\u00fcmk\u00fcnd\u00fcr. Bunun zihinsel durumlar i\u00e7in nomolojik olarak m\u00fcmk\u00fcn olup olmad\u0131\u011f\u0131 ise bana kal\u0131rsa ku\u015fkuludur, n\u00f6ral ikame d\u00fc\u015f\u00fcnce deneyine y\u00f6neltilen ele\u015ftiriler bunu \u00f6rnekler. Bizim ama\u00e7lar\u0131m\u0131z a\u00e7\u0131s\u0131ndan substrat esnekli\u011fini, substratlar aras\u0131nda de\u011fi\u015fmezlik gerektiren bir \u00e7oklu ger\u00e7ekle\u015febilirlik bi\u00e7imi olarak d\u00fc\u015f\u00fcnmek yeterlidir. Bu ilgin\u00e7 noktalar\u0131 g\u00fcndeme getiren hakeme m\u00fcte\u015fekkirim.<\/p>\n\n\n\n

[10]<\/a> Chalmers, kendi versiyonunda, buna aksonlar\u0131 ve di\u011fer n\u00f6ral ba\u011flant\u0131sall\u0131k unsurlar\u0131n\u0131 da ekler.<\/p>\n\n\n\n

[11]<\/a> Bu nokta yaln\u0131zca nitrik oksitle s\u0131n\u0131rl\u0131 de\u011fildir. Dopamin ve noradrenalin gibi monoaminerjik n\u00f6roiletkenler de geni\u015f sinirsel alanlarda etki g\u00f6sterir.<\/p>\n\n\n\n

[12]<\/a> Bu \u00f6rnek Ian Hinckfuss\u2019a aittir (Shiller, 2024). Tart\u0131\u015fma i\u00e7in bkz. Chalmers, 1996b; Dung & Kersten, 2024.<\/p>\n\n\n\n

[13]<\/a> Bu (nomolojik) itirazlar, Chalmers\u2019\u0131n bir insan \u00f6znenin kontrol\u00fcn\u00fcn ger\u00e7ek biyolojik beyin ile (varsay\u0131lan) i\u015flevsel kopyas\u0131 [duplicate<\/em>] aras\u0131nda gidip geldi\u011fi \u201cdans eden nitelceler\u201d [dancing qualia<\/em>] arg\u00fcman\u0131na da uygulan\u0131r (Chalmers, 1995; Godfrey-Smith, 2016).<\/p>\n\n\n\n

[14]<\/a> \u015e\u00f6yle denebilir: \u201cElbette zihinsel durumlar a\u00e7\u0131s\u0131ndan \u00f6nemli olan her \u015feyi yineleyen bir silikon beyin in\u015fa etmek zordur, ama varsayal\u0131m ki bunu yapabiliyoruz.\u201d Ne var ki, ger\u00e7ekle\u015fmesinin (nomolojik olarak) m\u00fcmk\u00fcn olmad\u0131\u011f\u0131na dair g\u00fc\u00e7l\u00fc nedenler bulunan durumlar\u0131 \u201cs\u0131rf hayal ederek\u201d i\u00e7inde ya\u015fad\u0131\u011f\u0131m\u0131z d\u00fcnyaya dair yararl\u0131 sonu\u00e7lar \u00e7\u0131karmak g\u00fc\u00e7t\u00fcr. Bu, genel olarak tasavvur edilebilirlik [conceivability<\/em>] arg\u00fcmanlar\u0131n\u0131n bir zay\u0131fl\u0131\u011f\u0131d\u0131r. Me\u015fhur ya da k\u00f6t\u00fc \u015f\u00f6hretli zombi arg\u00fcman\u0131 da benzer bir sorun ta\u015f\u0131r (Chalmers, 1996a; Seth, 2021).<\/p>\n\n\n\n

[15]<\/a> Kleiner (2024), ispat temelli bir y\u00f6ntemle benzer bir sonuca ula\u015f\u0131r. Bkz. ayr\u0131ca Kleiner & Ludwig (2023).<\/p>\n\n\n\n

[16]<\/a> Antikythera d\u00fczene\u011fi, gezegen hareketlerini \u00f6ng\u00f6rmek i\u00e7in kullan\u0131lan karma\u015f\u0131k bir mekanizmad\u0131r ve bir analog bilgisayar olarak yorumlanabilir. Bkz. https:\/\/www.wired.com\/story\/unbelievable-zombie-comeback-analog-computing\/<\/a><\/p>\n\n\n\n

[17]<\/a> Bkz. https:\/\/corticallabs.com\/cl1.html<\/a><\/p>\n\n\n\n

[18]<\/a> Adrian Thompson\u2019\u0131n, \u00e7e\u015fitli i\u015flevleri yerine getirmeleri i\u00e7in alan programlanabilir kap\u0131 dizileri (FPGA) yongalar\u0131n\u0131 genetik algoritmalarla optimize etmeye y\u00f6nelik erken \u00e7al\u0131\u015fmalar\u0131 buna \u00e7arp\u0131c\u0131 bir \u00f6rnek sunar. Thompson, baz\u0131 yap\u0131land\u0131rmalar\u0131n, FPGA bile\u015fenleri aras\u0131ndaki uzamsal ili\u015fkileri ve yongan\u0131n g\u00fc\u00e7 kayna\u011f\u0131n\u0131 \u00f6rt\u00fck bi\u00e7imde kulland\u0131\u011f\u0131n\u0131 bulmu\u015ftur. B\u00f6ylece yongan\u0131n i\u015flevi, kesin uzamsal yerle\u015fimine ba\u011fl\u0131 hale gelmi\u015f ve \u00e7oklu ger\u00e7ekle\u015febilirli\u011fi ciddi bi\u00e7imde k\u0131s\u0131tlam\u0131\u015ft\u0131r. Di\u011fer FPGA yap\u0131land\u0131rmalar\u0131n\u0131n ortam s\u0131cakl\u0131\u011f\u0131na ba\u011f\u0131ml\u0131 oldu\u011fu anla\u015f\u0131lm\u0131\u015f, Thompson bu k\u0131s\u0131t\u0131 farkl\u0131 s\u0131cakl\u0131klara ayarlanm\u0131\u015f buzdolaplar\u0131nda FPGA\u2019lar\u0131n\u0131 optimize ederek a\u015fmaya \u00e7al\u0131\u015fm\u0131\u015ft\u0131r (Thompson, 1998).<\/p>\n\n\n\n

[19]<\/a> Buradaki fikir \u015fudur: Turing i\u015flemlemesinin dayatt\u0131\u011f\u0131 kat\u0131 yaz\u0131l\u0131m\/donan\u0131m ayr\u0131m\u0131, yaz\u0131l\u0131mla ger\u00e7ekle\u015ftirilen i\u015flevler hakk\u0131nda g\u00f6zlemciye ba\u011f\u0131ml\u0131l\u0131k ima eder. \u00c7\u00fcnk\u00fc s\u00f6zdizimsel i\u015flemlere anlam (semantik) atfetmek i\u00e7in bir g\u00f6zlemciye ihtiya\u00e7 duyulur. \u0130\u015flemlemeselci i\u015flevselcilik bu durumda bilincin de g\u00f6zlemciye ba\u011f\u0131ml\u0131 oldu\u011funu ima edecektir, oysa bilin\u00e7 b\u00f6yle de\u011fildir. Bu arg\u00fcman\u0131 hat\u0131rlatt\u0131klar\u0131 i\u00e7in Simon Bowes ve Ray Tallis\u2019e te\u015fekk\u00fcrler.<\/p>\n\n\n\n

[20]<\/a> N\u00f6ral etkinli\u011fin dinamik y\u00f6nlerini bilin\u00e7li YZ\u2019nin ikna edici bulunmamas\u0131yla ili\u015fkilendiren bir arg\u00fcman i\u00e7in bkz. Kleiner & Ludwig, 2023.<\/p>\n\n\n\n

[21]<\/a> E\u011fer X bir Turing-d\u0131\u015f\u0131 i\u015flemleme s\u00fcreci ise, bir (Turing) sim\u00fclasyonu toplam girdi-\u00e7\u0131kt\u0131 e\u015fle\u015fmesi korunmu\u015f olsa bile tipik olarak farkl\u0131 i\u015flemlemeleri i\u00e7erecektir. Bunun bilince etkisinin olup olmad\u0131\u011f\u0131, i\u015flemlemeselci i\u015flevselcili\u011fin yorumuna do\u011frudan ba\u011fl\u0131d\u0131r: Tam olarak hangi i\u015flemlemenin y\u00fcr\u00fct\u00fcld\u00fc\u011f\u00fc m\u00fc \u00f6nemlidir, yoksa sadece toplam i\u015flemlemesel davran\u0131\u015f m\u0131? Burada farkl\u0131 g\u00f6r\u00fc\u015flere yer vard\u0131r.<\/p>\n\n\n\n

[22]<\/a> \u201cSim\u00fcle edilmi\u015f bir f\u0131rt\u0131na, sim\u00fclasyon i\u00e7indeki sim\u00fcle edilmi\u015f bir \u00f6zneye \u0131slak hissettirmez mi?\u201d Bu itiraz, sim\u00fcle edilmi\u015f bir f\u0131rt\u0131nay\u0131 sim\u00fclasyon oldu\u011funu fark edip etmedi\u011fimizin, f\u0131rt\u0131nay\u0131 \u201ci\u00e7eriden\u201d mi yoksa \u201cd\u0131\u015far\u0131dan\u201d m\u0131 deneyimledi\u011fimize ba\u011fl\u0131 oldu\u011funu \u00f6ne s\u00fcrer (Chalmers, 2022; Hofstadter, 1981). \u0130tiraz ge\u00e7erli de\u011fildir, \u00e7\u00fcnk\u00fc deneyim i\u00e7in i\u015flemlemenin yeterli oldu\u011funu pe\u015finen varsayar. Bkz. ayr\u0131ca Wiese, 2024.<\/p>\n\n\n\n

[23]<\/a> Kime atfedilece\u011fi tart\u0131\u015fmal\u0131 bir vecizede \u015f\u00f6yle denmi\u015ftir: \u201cBilgisayarlar asl\u0131nda \u2018satranc\u0131 oynamazlar\u2019, \u2018satranc\u0131n tarihini\u2019 oynarlar.\u201d<\/p>\n\n\n\n

[24]<\/a> Bayes\u00e7i anlamda \u201cinan\u00e7\u201d, (ki\u015fi-d\u00fczeyinde) psikolojik bir inanc\u0131 ima etmez. Bayes\u00e7i inan\u00e7lar, genellikle bir ortalama ve de\u011fi\u015fke (varyans) ya da bir kesinlikle karakterize edilen olas\u0131l\u0131k da\u011f\u0131l\u0131mlar\u0131d\u0131r.<\/p>\n\n\n\n

[25]<\/a> \u00d6ng\u00f6r\u00fcc\u00fc i\u015fleme genel olarak \u015fu konuda agnostiktir: (Bilin\u00e7li) alg\u0131sal i\u00e7erikler ger\u00e7ekten alg\u0131sal \u201cen iyi tahminler\u201d taraf\u0131ndan m\u0131 olu\u015fturulur, yoksa yaln\u0131zca bu \u015fekilde mi en iyi modellenir? Bkz. B\u00f6l\u00fcm 4.2 ve Andrews, 2021; Clark, 2023; Friston, 2010; Hohwy & Seth, 2020.<\/p>\n\n\n\n

[26]<\/a> Allostaz, homeostaz\u0131n genelle\u015ftirilmesidir ve de\u011fi\u015fim boyunca istikrar\u0131 vurgular. \u00d6rne\u011fin aya\u011fa kalkt\u0131\u011f\u0131m\u0131zda bay\u0131lmam\u0131z\u0131 \u00f6nlemek i\u00e7in kan bas\u0131nc\u0131m\u0131z ge\u00e7ici olarak y\u00fckselir (Sterling, 2012).<\/p>\n\n\n\n

[27]<\/a> Burada bir ba\u015fka a\u00e7\u0131, bilin\u00e7li deneyimin zamansal yanlar\u0131n\u0131n, Husserlci \u201cprotensiyon\u201d ve \u201cretensiyon\u201dla ili\u015fkilendirilen \u201ckal\u0131nl\u0131\u011f\u0131n\u201d [thickness<\/em>] (Husserl, 1982 [1913]), hele ki \u00e7\u0131kar\u0131m zamana yay\u0131lan serbest enerji minimizasyonu olarak kavrand\u0131\u011f\u0131 takdirde, \u00f6ng\u00f6r\u00fcc\u00fc \u00e7\u0131kar\u0131m\u0131n zamansal uzan\u0131m\u0131yla ili\u015fkili olabilece\u011fidir. Bkz. B\u00f6l\u00fcm 4.2 ve Friston, 2018; Kiverstein, 2020.<\/p>\n\n\n\n

[28]<\/a> Ayr\u0131ca bkz. Thompson & Varela, 2001. Asgari fenomenal deneyime farkl\u0131 bir yakla\u015f\u0131m i\u00e7in bkz. Metzinger, 2024.<\/p>\n\n\n\n

[29]<\/a> \u0130nsan tasar\u0131m\u0131 otopoietik sistem \u00f6rnekleri var m\u0131d\u0131r? Sentetik biyolojik sistemler bu duruma en \u00e7ok yakla\u015fanlard\u0131r. Ancak bug\u00fcne dek, (otopoietik) bile\u015fen \u00f6\u011felerinin, yani h\u00fccrelerin, en iyi ihtimalle insanlar taraf\u0131ndan de\u011fi\u015ftirildi\u011fi, b\u00fct\u00fcn\u00fcyle s\u0131f\u0131rdan (de novo<\/em>) yarat\u0131lmad\u0131\u011f\u0131 ileri s\u00fcr\u00fclebilir.<\/p>\n\n\n\n

[30]<\/a> Bu t\u00fcr s\u00fcreklilik anlay\u0131\u015flar\u0131, g\u00fc\u00e7l\u00fc form\u00fclasyonlar\u0131nda \u201czihin kelimenin tam anlam\u0131yla ya\u015fam gibidir<\/em>\u201d diyen \u201czihin\u2013ya\u015fam\u201d s\u00fcreklili\u011fi tezini \u00e7a\u011fr\u0131\u015ft\u0131r\u0131r (Godfrey-Smith, 1996, s. 320, vurgu yazara ait). Ayr\u0131ca bkz. Thompson, 2007. Zihin\u2013ya\u015fam s\u00fcreklili\u011fi, serbest enerji ilkesi ve otopoiesis \u00fczerine bir tart\u0131\u015fma i\u00e7in bkz. Kirchhoff & Froese, 2017.<\/p>\n\n\n\n

[31]<\/a> Bu yolculu\u011fun daha yava\u015f bir versiyonu i\u00e7in bkz. Seth, 2021, bl. 10; ili\u015fkili anlat\u0131lar i\u00e7in bkz. Kiverstein, 2020; Solms, 2021; serbest enerji ilkesi \u00fczerine daha fazlas\u0131 i\u00e7in bkz. Bogacz, 2017; Buckley ve di\u011ferleri, 2017; Friston, 2010; Nave, 2025; Parr ve di\u011ferleri, 2022.<\/p>\n\n\n\n

[32]<\/a> Bunu anlaman\u0131n bir ba\u015fka yolu da \u015fudur: Ya\u015fayan sistemler, dalgalanma teoreminin<\/em> [fluctuation theorem<\/em>] sundu\u011fu olanaklardan yararlan\u0131r. Bu teorem, yal\u0131t\u0131lm\u0131\u015f sistemlerin entropide ge\u00e7ici azalmalar ya\u015fama olas\u0131l\u0131\u011f\u0131n\u0131 nicelendirir (Evans ve di\u011ferleri, 1993).<\/p>\n\n\n\n

[33]<\/a> Daha kesin tan\u0131mla serbest enerji, verili bir modelde bir \u015feyin (istatistiksel olarak) ne \u00f6l\u00e7\u00fcde beklenmedik oldu\u011funu belirleyen s\u00fcrpriz<\/em> [surprisal<\/em>] i\u00e7in bir \u00fcst s\u0131n\u0131r sa\u011flar. Dengeden uzak kararl\u0131 durum varsay\u0131mlar\u0131 alt\u0131nda s\u00fcrprizin uzun d\u00f6nem ortalamas\u0131 entropidir. Bkz. Buckley ve di\u011ferleri, 2017; Parr ve di\u011ferleri, 2022.<\/p>\n\n\n\n

[34]<\/a> Ba\u015fl\u0131ca varsay\u0131m, ilgili olas\u0131l\u0131k da\u011f\u0131l\u0131mlar\u0131n\u0131n Gauss da\u011f\u0131l\u0131mlar\u0131 olarak parametrele\u015ftirilebilmesidir. Bu (Laplace) varsay\u0131m(\u0131) alt\u0131nda, \u00f6ng\u00f6r\u00fc hatalar\u0131 serbest enerjilerin kendileriyle de\u011fil, serbest enerji gradyanlar\u0131yla ili\u015fkilendirilir. Bkz. Friston ve Kiebel, 2009; Buckley ve di\u011ferleri, 2017; Parr ve di\u011ferleri, 2022.<\/p>\n\n\n\n

[35]<\/a> Bu sonu\u00e7, (varyasyonel) serbest enerjinin Bayes\u00e7i model kan\u0131t\u0131 i\u00e7in bir alt s\u0131n\u0131r sa\u011flamas\u0131ndan kaynaklan\u0131r; bu alt s\u0131n\u0131r kan\u0131t alt s\u0131n\u0131r\u0131 [evidence lower bound, ELBO<\/em>] olarak adland\u0131r\u0131l\u0131r. Serbest enerjinin en aza indirilmesi, ELBO\u2019nun azamiye \u00e7\u0131kar\u0131lmas\u0131yla e\u015fde\u011ferdir. Bkz. Parr ve di\u011ferleri, 2022.<\/p>\n\n\n\n

[36]<\/a> Lane, bu \u00e7\u0131kt\u0131n\u0131n zar boyunca elektrostatik ve elektromanyetik potansiyellerin oran\u0131 taraf\u0131ndan sa\u011fland\u0131\u011f\u0131n\u0131 \u00f6ne s\u00fcrer (Lane, 2022). Ayr\u0131ca bkz. (Alexandre, 2010; Egbert ve di\u011ferleri, 2010).<\/p>\n\n\n\n

[37]<\/a> Bu ifade i\u00e7in David Chalmers\u2019a sayg\u0131lar\u0131m\u0131 sunuyorum. Wanja Wiese yak\u0131n zamanda benzer bir g\u00f6r\u00fc\u015f \u00f6nererek buna \u201czay\u0131f i\u015flemlemeselci i\u015flevselcilik\u201d ad\u0131n\u0131 vermi\u015ftir (Wiese, 2024).<\/p>\n\n\n\n

[38]<\/a> Kleiner (2024), i\u015flemlemeselci i\u015flevselcilik do\u011fruysa ilgili i\u015flemlemelerin \u00f6l\u00fcml\u00fc olmas\u0131 gerekti\u011fini, dolay\u0131s\u0131yla standart (\u00f6l\u00fcms\u00fcz) YZ\u2019nin bilin\u00e7li olamayaca\u011f\u0131n\u0131 savunur. Ama bu durumda, i\u015flemlemeselci i\u015flevselcili\u011fin baz\u0131 yorumlar\u0131na g\u00f6re, \u00f6l\u00fcms\u00fcz i\u015flemleme y\u00f6ntemleri kullan\u0131larak ilgili \u00f6l\u00fcml\u00fc i\u015flemlemelerin yeterince iyi bir sim\u00fclasyonunun yap\u0131lmas\u0131 da bilinci ger\u00e7ekle\u015ftirmelidir. Bu g\u00f6r\u00fcn\u00fcrde bir \u00e7eli\u015fki olu\u015fturur. Bir \u00e7\u00f6z\u00fcm, belki de i\u015flemlemeler farkl\u0131 oldu\u011fu i\u00e7in, \u00f6l\u00fcml\u00fc i\u015flemlemelerin sim\u00fclasyonunun yeterli olmad\u0131\u011f\u0131n\u0131 s\u00f6ylemektir (Bkz. 21. not). \u00c7eli\u015fki, i\u015flemlemeselci i\u015flevselcilik yanl\u0131\u015fsa da \u00e7\u00f6z\u00fclecektir. Ve e\u011fer i\u015flemlemeselci i\u015flevselcilik yanl\u0131\u015fsa, bilin\u00e7li YZ zaten m\u00fcmk\u00fcn de\u011fildir.<\/p>\n\n\n\n

[39]<\/a> Ya\u015fam ile bilinci serbest enerji ilkesi \u00fczerinden ili\u015fkilendiren alternatif bir anlat\u0131m i\u00e7in bkz. Wiese, 2024. Buradaki temel fikir kabaca \u015fudur: bilin\u00e7, bir sistemin i\u015flemlemesel dinamikleri ile fiziksel dinamikleri aras\u0131nda bir e\u015fle\u015fmeyi gerektirir.<\/p>\n\n\n\n

[40]<\/a> Daha \u00e7ok LLM\u2019lerle ili\u015fkili ba\u015fka bir olas\u0131l\u0131k da, bilin\u00e7le ilgili i\u00e7erimlerinden ba\u011f\u0131ms\u0131z olarak, sahici bir anlama<\/em> i\u00e7in bedenlenmi\u015flik ve g\u00f6m\u00fcl\u00fcl\u00fc\u011f\u00fcn gerekli olabilece\u011fidir.<\/p>\n\n\n\n

[41]<\/a> Bir kez daha vurgulayal\u0131m, Turing Testi zek\u00e2y\u0131 s\u0131nar, bilinci de\u011fil. Bilince y\u00f6nelik benzer bir test fikri, Alex Garland\u2019\u0131n Ex Machina<\/em> filmi onuruna Garland Testi olarak adland\u0131r\u0131lm\u0131\u015ft\u0131r (Shanahan, 2016). Ayr\u0131ca bkz. Seth, 2021, bl. 13. YZ\u2019de bilin\u00e7 i\u00e7in bir test \u00f6nerisi i\u00e7in bkz. Schneider, 2019.<\/p>\n\n\n\n

[42]<\/a> Peter Godfrey-Smith\u2019in \u201cdar duyumsamac\u0131l\u0131k\u201d [sentientism<\/em>] ile \u201cgeni\u015f duyumsamac\u0131l\u0131k\u201d ayr\u0131m\u0131 i\u00e7in bkz. https:\/\/metazoan.net\/108-whitehead-lectures\/<\/a>. Farisco ve di\u011ferleri (2024), varsay\u0131lan ger\u00e7ek yapay bilin\u00e7 durumlar\u0131n\u0131n nitelik\u00e7e insan bi\u00e7iminden farkl\u0131 olabilece\u011fine dikkat \u00e7eker.<\/p>\n\n\n\n

[43]<\/a> Meseleyi daha da karma\u015f\u0131kla\u015ft\u0131racak bi\u00e7imde Long ve \u00e7al\u0131\u015fma arkada\u015flar\u0131, bilin\u00e7siz YZ\u2019nin \u00e7\u0131karlar\u0131n\u0131n ya da \u201ctercihlerinin\u201d engellenmesinin bile etik bak\u0131mdan sorunlu olabilece\u011fi olas\u0131l\u0131\u011f\u0131n\u0131 g\u00fcndeme getirir (Long ve di\u011ferleri, 2024). Yoshua Bengio\u2019nun, YZ i\u00e7in uygun bir hedefin bilge ama kendi \u00e7\u0131karlar\u0131 ya da eyleyicili\u011fi olmayan bir \u201ck\u00e2hin\u201d olmas\u0131 gerekti\u011fi y\u00f6n\u00fcndeki \u00f6nerisi bu sorunlar\u0131 bir \u00f6l\u00e7\u00fcde yat\u0131\u015ft\u0131rabilir. Bkz. https:\/\/yoshuabengio.org\/2023\/05\/07\/ai-scientists-safe-and-useful-ai\/<\/a>.<\/p>\n\n\n\n

[44]<\/a> \u201cSentetik fenomenoloji\u201d i\u00e7erebilecek her t\u00fcrl\u00fc ara\u015ft\u0131rma i\u00e7in moratoryum \u00e7a\u011fr\u0131s\u0131 yapanlar da vard\u0131r (Metzinger, 2021).<\/p>\n\n\n\n

[45]<\/a> Bir ba\u015fka olas\u0131l\u0131k, LLM\u2019ler gibi YZ sistemlerinin o denli iyi h\u00e2le gelmesidir ki art\u0131k insans\u0131 g\u00f6r\u00fcnmezler. Ger\u00e7ek olamayacak kadar iyi olduklar\u0131 i\u00e7in tekinsiz vadiden [uncanny valley<\/em>] \u00e7\u0131k\u0131p b\u00fct\u00fcn\u00fcyle ba\u015fka bir yere y\u00f6nelebilirler. 2025\u2019in ba\u015flar\u0131nda DeepSeek-R1 ile etkile\u015fime girdi\u011fimde buna benzer bir deneyim ya\u015fad\u0131m.<\/p>\n\n\n\n

[46]<\/a> https:\/\/www.technologyreview.com\/2024\/05\/07\/1092116\/deepfakes-dead-chinese-business-grief\/<\/a><\/p>\n\n\n\n

[47]<\/a> https:\/\/www.ft.com\/content\/96187a7a-fce5-11e6-96f8-3700c5664d30<\/a><\/p>\n\n\n\n

[48]<\/a> Makineleri insanbi\u00e7imcile\u015ftirmenin ve zihinleri \u201cbilgisayarla\u015ft\u0131rman\u0131n\u201d zararl\u0131 sonu\u00e7lar\u0131 \u00fczerine bkz. Floridi & Nobre, 2024.<\/p>\n\n\n\n

[49]<\/a> Bu yeni geli\u015fmeler, biyolojik metaforlar\u0131n kar\u015f\u0131l\u0131kl\u0131 [reciprocal<\/em>] do\u011fas\u0131n\u0131 da \u00f6ne \u00e7\u0131kar\u0131r. Biyolojik olgular\u0131 anlama tarz\u0131m\u0131z makineler ve bilgisayarlar \u00fczerinden \u015fekillenmi\u015f, buna kar\u015f\u0131l\u0131k biyoloji ve sinirbilimden gelen i\u00e7g\u00f6r\u00fcler de makineleri ve bilgisayarlar\u0131 anlama tarz\u0131m\u0131z\u0131 etkilemi\u015ftir (Barwich & Rodriguez, 2024).<\/p>\n\n\n\n

[50]<\/a> https:\/\/podcasts.apple.com\/gb\/podcast\/frankenstein-with-anil-seth-and-fiona-sampson\/id1549179379?i=1000637016221<\/a><\/p>\n\n\n\n

<\/div>\n\n\n\n

Referanslar<\/strong><\/p>\n\n\n\n

<\/div>\n\n\n\n

Aguilera, M., Millidge, B., Tschantz, A., & Buckley, C. L. (2022). How particular is the physics of
the free energy principle? Phys Life Rev, 40, 24-50. https:\/\/doi.org\/10.1016\/j.plrev.2021.11.001<\/p>\n\n\n\n

Alais, D., Xu, Y., Wardle, S. G., & Taubert, J. (2021). A shared mechanism for facial expression in
human faces and face pareidolia. Proc Biol Sci, 288(1954), 20210966. https:\/\/doi.org\/10.1098\/rspb.2021.0966<\/p>\n\n\n\n

Albantakis, L., Barbosa, L., Findlay, G., Grasso, M., Haun, A. M., Marshall, W., Mayner, W. G. P., Zaeemzadeh, A., Boly, M., Juel, B. E., Sasai, S., Fujii, K., David, I., Hendren, J., Lang, J. P., & Tononi, G. (2023). Integrated information theory (IIT) 4.0: Formulating the properties of phenomenal existence in physical terms. PLoS Comput Biol, 19(10), e1011465. https:\/\/doi.org\/10.1371\/journal.pcbi.1011465
<\/p>\n\n\n\n

Alexandre, G. (2010). Coupling metabolism and chemotaxis-dependent behaviours by energy taxis receptors. Microbiology (Reading), 156(Pt 8), 2283-2293. https:\/\/doi.org\/10.1099\/mic.0.039214-0<\/p>\n\n\n\n


Anderson, N. G., & Piccinini, G. (2024). The physical signature of computation: A robust mapping account. Oxford University Press.<\/p>\n\n\n\n


Anderson, N. M., & Piccinini, G. (2024). The physical signature of computation. Oxford University Press.<\/p>\n\n\n\n


Andrews, M. (2021). The math is not the territory: navigating the free energy principle. Biology and Philosophy, 36, 30.<\/p>\n\n\n\n


Aru, J., Larkum, M. E., & Shine, J. M. (2023). The feasibility of artificial consciousness through the lens of neuroscience. Trends Neurosci, 46(12), 1008-1017.
https:\/\/doi.org\/10.1016\/j.tins.2023.09.009<\/p>\n\n\n\n


Ball, P. (2023). How life works: A user\u2019s guide to the new biology. University of Chicago Press.<\/p>\n\n\n\n


Barnett, L., & Seth, A. K. (2023). Dynamical independence: Discovering emergent macroscopic
processes in complex dynamical systems. Phys Rev E, 108(1-1), 014304.
https:\/\/doi.org\/10.1103\/PhysRevE.108.014304<\/p>\n\n\n\n


Barrett, L. F., & Simmons, W. K. (2015). Interoceptive predictions in the brain. Nat Rev Neurosci,
16(7), 419-429. https:\/\/doi.org\/10.1038\/nrn3950<\/p>\n\n\n\n


Barwich, A.-S., & Rodriguez, M. J. (2024). Rage against the what? The machine metaphor in
biology. Biology and Philosophy, 39, 14.<\/p>\n\n\n\n


Bayne, T., Seth, A. K., & Massimini, M. (2020). Are There Islands of Awareness? Trends Neurosci,
43(1), 6-16. https:\/\/doi.org\/10.1016\/j.tins.2019.11.003<\/p>\n\n\n\n


Bayne, T., Seth, A. K., Massimini, M., Shepherd, J., Cleeremans, A., Fleming, S. M., Malach, R., Mattingley, J. B., Menon, D. K., Owen, A. M., Peters, M. A. K., Razi, A., & Mudrik, L. (2024). Tests for consciousness in humans and beyond. Trends Cogn Sci, 28(5), 454-466.
https:\/\/doi.org\/10.1016\/j.tics.2024.01.010<\/p>\n\n\n\n


Bechtel, W., & Mundale, J. (1999). Multiple realizability revisited: Linking cognitive and neural
states. Philoosophy of Science, LXVI, 175-207.<\/p>\n\n\n\n

Biehl, M., Pollock, F. A., & Kanai, R. (2021). A Technical Critique of Some Parts of the Free Energy Principle. Entropy (Basel), 23(3). https:\/\/doi.org\/10.3390\/e23030293<\/p>\n\n\n\n


Birch, J. (2024). The edge of sentience. Oxford University Press.<\/p>\n\n\n\n


Block, N. (2019). Fading qualia: A response to Michael Tye. In P. A. & D. Stoljar (Eds.), Blockheads! Essays on Ned Block\u2019s philosophy of mind and consciousness. MIT Press.<\/p>\n\n\n\n


Blum, L., & Blum, M. (2024). AI Consciousness is Inevitable: A Theoretical Computer Science Perspective. ArXiv, arXiv:2403.17101. https:\/\/doi.org\/10.48550\/arXiv.2403.17101<\/p>\n\n\n\n


Boden, M. (2008). Mind as machine: A history of cognitive science. Oxford University Press.<\/p>\n\n\n\n


Bogacz, R. (2017). A tutorial on the free-energy framework for modelling perception and learning. J Math Psychol, 76(Pt B), 198-211. https:\/\/doi.org\/10.1016\/j.jmp.2015.11.003<\/p>\n\n\n\n


Bostrom, N. (2003). Are you living in a computer simulation? Philosophical Quarterly, 53(11),
243-255.<\/p>\n\n\n\n


Brown, R., Lau, H., & LeDoux, J. E. (2019). Understanding the Higher-Order Approach to
Consciousness. Trends Cogn Sci, 23(9), 754-768.
https:\/\/doi.org\/10.1016\/j.tics.2019.06.009<\/p>\n\n\n\n


Bruineberg, J., Dolega, K., Dewhurst, J., & Baltieri, M. (2021). The Emperor’s New Markov Blankets. Behav Brain Sci, 45, e183. https:\/\/doi.org\/10.1017\/S0140525X21002351<\/p>\n\n\n\n


Buckley, C., Kim, C.-S., McGregor, S., & Seth, A. K. (2017). The free energy principle for action and
perception: A mathematical review. Journal of Mathematical Psychology, 81, 55-79.<\/p>\n\n\n\n


Butlin, P., Long, R., Elmoznino, E., & Bengio, Y. (2023). Consciousness in Artificial Intelligence:
Insights from the Science of Consciousness. Https:\/\/doi.org\/10.48550\/arXiv.2308.08708<\/p>\n\n\n\n


Cantwell Smith, B. (2002). The Foundations of Computing. In M. Scheutz (Ed.), Computationalism: New Directions. Cambridge, MA.<\/p>\n\n\n\n


Cao, R. (2022). Multiple realizability and the spirit of functionalism. Synthese, 200, 506.<\/p>\n\n\n\n


Chalmers, D. (1995). Absent qualia, fading qualia, dancing qualia. In T. Metzinger (Ed.), Conscious
Experience (pp. 309-328). Ferdinand Schoningh.<\/p>\n\n\n\n


Chalmers, D. J. (1994). On implementing a computation. Minds and Machines, 4, 391-402.<\/p>\n\n\n\n


Chalmers, D. J. (1996a). The conscious mind: In search of a fundamental theory. Oxford
University Press.<\/p>\n\n\n\n


Chalmers, D. J. (1996b). Does a rock implement every finite-state automaton? Synthese, 108(3),
309-333.<\/p>\n\n\n\n


Chalmers, D. J. (2003). Consciousness and its place in nature. In S. Stich & T. A. Warfield (Eds.),
The Blackwell Guide to the Philosophy of Mind. Blackwell Publlishing Ltd.<\/p>\n\n\n\n


Chalmers, D. J. (2022). Reality+: Virtual worlds and the problems of philosophy. W. W. Norton &
Company.<\/p>\n\n\n\n


Chalmers, D. J. (2023). Could a Large Language Model be Conscious? https:\/\/arxiv.org\/abs\/2303.07103<\/p>\n\n\n\n


Chintaluri, C., & Vogels, T. P. (2023). Metabolically regulated spiking could serve neuronal energy
homeostasis and protect from reactive oxygen species. Proc Natl Acad Sci U S A, 120(48),
e2306525120. https:\/\/doi.org\/10.1073\/pnas.2306525120<\/p>\n\n\n\n

Chrisley, R. (2020). A Human-Centered Approach to AI Ethics: A Perspective from Cognitive
Science. In M. Dubber, F. Pasquale, & S. Das (Eds.), The Oxford Handbook of Ethics of AI
(pp. 463-474). Oxford University Press.<\/p>\n\n\n\n


Churchland, P. S. (1996). The hornswoggle problem. Journal of Consciousness Studies, 3(5-6),
402-408.<\/p>\n\n\n\n


Clark, A. (2013). Whatever next? Predictive brains, situated agents, and the future of cognitive
science. Behav Brain Sci, 36(3), 181-204. https:\/\/doi.org\/10.1017\/S0140525X12000477<\/p>\n\n\n\n


Clark, A. (2019). Consciousness as generative entanglement. The Journal of Philosophy, 116(12),
645-662.<\/p>\n\n\n\n


Clark, A. (2023). The experience machine: How our minds predict and shape reality. Allen Lane.<\/p>\n\n\n\n


Cobb, M. (2020). The idea of the brain: A history. Profile Books.<\/p>\n\n\n\n


Cole, D. (2023). The Chinese Room Argument. In E. N. Zalta & U. Nodelman (Eds.), The Stanford
Encyclopedia of Philosophy. https:\/\/plato.stanford.edu\/archives\/sum2023\/entries\/chinese-room\/<\/p>\n\n\n\n


Collell, G., & Fauquet, J. (2015). Brain activity and cognition: a connection from thermodynamics
and information theory. Frontiers in Psychology, 6, 818.<\/p>\n\n\n\n


Colombatto, C., & Fleming, S. M. (2023). Folk psychological attribuitions of consciousness to
large language models. https:\/\/doi.org\/ https:\/\/doi.org\/10.31234\/osf.io\/5cnrv<\/p>\n\n\n\n


Conant, R., & Ashby, W. R. (1970). Every good regulator of a system must be a model of that
system. International Journal of Systems Science, 1(2), 89-97.<\/p>\n\n\n\n


Copeland, J. B. (2024). The Church-Turing thesis. In E. N. Zalta & U. Nodelman (Eds.), The
Stanford Encyclopedia of Philosophy. Metaphysics Research Lab, Stanford University.<\/p>\n\n\n\n


Cosmelli, D., Lachaux, J.-P., & Thompson, E. (2007). Neurodynamics of Consciousness. In P. D.
Zelazo, M. Moscovitch, & E. Thompson (Eds.), The Cambridge Handbook of Consciousness
(pp. 731-775). Cambridge University Press.<\/p>\n\n\n\n


Crick, F., & Koch, C. (1990). Towards a neurobiological theory of consciousness. Seminars in the
Neurosciences, 2, 263-275.<\/p>\n\n\n\n


Damasio, A. (2000). The feeling of what happens: Body and emotion in the making of
consciousness. Harvest Books.<\/p>\n\n\n\n


Dehaene, S., Lau, H., & Kouider, S. (2017). What is consciousness, and could machines have it?
Science, 358(6362), 486-492. https:\/\/doi.org\/10.1126\/science.aan8871<\/p>\n\n\n\n


Dennett, D. (1987). The intentional stance. MIT Press.<\/p>\n\n\n\n


Di Paolo, E., Thompson, E., & Beer, R. D. (2022). Laying down a forking path: Tensions between
enaction and the free energy principle. Philosophy and the Mind Sciences, 3.<\/p>\n\n\n\n


Douglas, R., Mahowald, M., & Mead, C. (1995). Neuromorphic analogue VLSI. Annual Review of
Neuroscience, 18, 255-281.<\/p>\n\n\n\n


Dung, L., & Kersten, L. (2024). Implementing artitifical consciousness. Mind and Language.
https:\/\/doi.org\/10.1111\/mila.12532<\/p>\n\n\n\n


Dupuy, J.-P. (2009). On the origins of cognitive science: The mechanization of mind (2nd ed.). MIT
Press.<\/p>\n\n\n\n


Ebrahimkhani, M. R., & Levin, M. (2021). Synthetic living machines: A new window on life.
iScience, 24(5), 102505. https:\/\/doi.org\/10.1016\/j.isci.2021.102505<\/p>\n\n\n\n


Edelman, G. M. (1989). The remembered present. Basic Books.<\/p>\n\n\n\n

Egbert, M. D., Barandiaran, X. E., & Di Paolo, E. A. (2010). A minimal model of metabolism-based
chemotaxis. PLoS Comput Biol, 6(12), e1001004. https:\/\/doi.org\/10.1371\/journal.pcbi.1001004<\/p>\n\n\n\n


Evans, D. J., Cohen, E. G., & Morriss, G. P. (1993). Probability of second law violations in shearing
steady states. Phys Rev Lett, 71(15), 2401-2404. https:\/\/doi.org\/10.1103\/PhysRevLett.71.2401<\/p>\n\n\n\n


Evers, K., Farisco, M., Chatila, R., Earp, B. D., Freire, I. T., Hamker, F., Nemeth, E., Verschure, P., &
Khamassi, M. (2025). Preliminaries to artificial consciousness: A multidimensional heuristic approach. Phys Life Rev, 52, 180-193. https:\/\/doi.org\/10.1016\/j.plrev.2025.01.002<\/p>\n\n\n\n


Farisco, M., Evers, K., & Changeux, J. P. (2024). Is artificial consciousness achievable? Lessons
from the human brain. Neural Netw, 180, 106714. https:\/\/doi.org\/10.1016\/j.neunet.2024.106714<\/p>\n\n\n\n


Fields, C., Goldstein, A., & Sandved-Smith, L. (2024). Making the Thermodynamic Cost of Active
Inference Explicit. Entropy (Basel), 26(8). https:\/\/doi.org\/10.3390\/e26080622<\/p>\n\n\n\n


Findlay, G., Marshall, W., Albantakis, L., David, I., Mayner, W. G. P., Koch, C., & Tononi, G. (2024).
Dissociating artificial intelligence from artificial consciousness. ArXiv, 2412, 04571 https:\/\/doi.org\/https:\/\/doi.org\/10.48550\/arXiv.2412.04571<\/p>\n\n\n\n


Fleming, S. M. (2020). Awareness as inference in a higher-order state space. Neurosci Conscious, 2020(1), niz020. https:\/\/doi.org\/10.1093\/nc\/niz020<\/p>\n\n\n\n


Fleming, S. M. (2021). Know thyself. Basic Books.<\/p>\n\n\n\n


Floridi, L. (2010). Information: A very short introduction. Oxford University Press.<\/p>\n\n\n\n


Floridi, L., & Nobre, A. C. (2024). Anthropomorphising Machines and Computerising Minds: The Crosswiring of Languages between Artificial Intelligence and Brain & Cognitive Sciences. Minds and Machines, 34. https:\/\/doi.org\/https:\/\/doi.org\/10.1007\/s11023-024-09670-4<\/p>\n\n\n\n


Freeman, W. J. (1999). Consciousness, intentionality, and causality. Journal of Consciousness Studies, 6(11-12), 143-172.<\/p>\n\n\n\n


Friston, K., & Kiebel, S. (2009). Predictive coding under the free-energy principle. Philos Trans R Soc Lond B Biol Sci, 364(1521), 1211-1221. http:\/\/www.ncbi.nlm.nih.gov\/entrez\/query.fcgicmd=Retrieve&db=PubMed&dopt=Citation&list_uids=19528002
<\/p>\n\n\n\n

Friston, K. J. (2010). The free-energy principle: a unified brain theory? Nat Rev Neurosci, 11(2), 127-138. https:\/\/doi.org\/nrn2787 [pii] <\/p>\n\n\n\n


Friston, K. J. (2013). Life as we know it. J R Soc Interface, 10(86), 20130475.
https:\/\/doi.org\/10.1098\/rsif.2013.0475<\/p>\n\n\n\n


Friston, K. J. (2018). Am I Self-Conscious? (Or Does Self-Organization Entail Self-Consciousness?).
Front Psychol, 9, 579. https:\/\/doi.org\/10.3389\/fpsyg.2018.00579<\/p>\n\n\n\n


Friston, K. J., FitzGerald, T., Rigoli, F., Schwartenbeck, P., & Pezzulo, G. (2017). Active Inference: A
Process Theory. Neural Comput, 29(1), 1-49. https:\/\/doi.org\/10.1162\/NECO_a_00912<\/p>\n\n\n\n


Froese, T. (2017). Life is precious because it is precarious: Individuality, mortality and the problem of meaning. In G. Dodig-Crnkovic & E. Giovagnoli (Eds.), Representation and reality in humans, other living organisms, and intelligent machines (pp. 33-50). Springer International Publishing.<\/p>\n\n\n\n

Godfrey-Smith, P. G. (1996). Spencer and Dewey on life and mind. In M. Boden (Ed.), The
philosophy of artificial life (pp. 314-331). Oxford University Press.<\/p>\n\n\n\n


Godfrey-Smith, P. G. (2016). Mind, matter, and metabolism. Journal of Philosophy, CXIII(10), 481-506.<\/p>\n\n\n\n

Goertzel, B., & Pennachin, C. (Eds.). (2007). Artificial general intelligence. Springer-Verlag.<\/p>\n\n\n\n


Gomez-Marin, A., & Seth, A. K. (2025). A science of consciousness beyond pseudo-science and
pseudo-consciousness. Nat Neurosci.<\/p>\n\n\n\n


Graziano, M. S. A. (2017). The attention schema theory: A foundation for engineering artificial
consciousness. Frontiers in Robotics and AI 4, 60. https:\/\/doi.org\/10.3389\/frobt.2017.00060<\/p>\n\n\n\n


Gumuskaya, G., Srivastava, P., Cooper, B. G., Lesser, H., Semegran, B., Garnier, S., & Levin, M. (2024). Motile Living Biobots Self-Construct from Adult Human Somatic Progenitor Seed Cells. Adv Sci (Weinh), 11(4), e2303575. https:\/\/doi.org\/10.1002\/advs.202303575<\/p>\n\n\n\n


Haeckel, E. (1892). Our monism: The principles of a consistent, unitary, world-view. The Monist, 2, 481-486.<\/p>\n\n\n\n


Hameroff, S., & Penrose, R. (2014). Consciousness in the universe: a review of the ‘Orch OR’ theory. Phys Life Rev, 11(1), 39-78. https:\/\/doi.org\/10.1016\/j.plrev.2013.08.002<\/p>\n\n\n\n


Haugeland, J. (1980). Programs, causal powers and intentionality. Behavioral and Brain Sciences, 3, 432-433.
<\/p>\n\n\n\n

Hinton, G. (2022). The forward-forward algorithm: some preliminary investigations. https:\/\/arxiv.org\/abs\/2212.13345
<\/p>\n\n\n\n

Hochstetter, J., Zhu, R., Loeffler, A., Diaz-Alvarez, A., Nakayama, T., & Kuncic, Z. (2021). Avalanches and edge-of-chaos learning in neuromorphic nanowire networks. Nat Commun, 12(1), 4008. https:\/\/doi.org\/10.1038\/s41467-021-24260-z
<\/p>\n\n\n\n

Hofstadter, D. (1981). A coffee-house conversation on the Turing test. Scientific American.
<\/p>\n\n\n\n

Hohwy, J. (2013). The Predictive Mind. Oxford University Press.<\/p>\n\n\n\n

Hohwy, J. (2014). The self-evidencing brain. Nous, 1-27. https:\/\/doi.org\/10.1111\/nous.12062
<\/p>\n\n\n\n

Hohwy, J., & Seth, A. K. (2020). Predictive processing as a systematic basis for identifying the neural correlates of consciousness. Philosophy and the Mind Sciences, 1(2), 3.
<\/p>\n\n\n\n

Husserl, E. (1982 [1913]). Ideas pertaining to a pure phenomenology and to a phenomenological philosophy – first book: General introduction to a pure phenomenology.
<\/p>\n\n\n\n

Jaeger, J., Riedl, A., Djedovic, A., Vervaeke, J., & Walsh, D. (2024). Naturalizing relevance realization: why agency and cognition are fundamentally not computational. Front Psychol, 15, 1362658 https:\/\/doi.org\/doi: 10.3389\/fpsyg.2024.1362658<\/p>\n\n\n\n

Jarzynski, C. (1997). Nonequilibrium equality for free energy differences. Phys Rev Lett, 78(14),
2690.<\/p>\n\n\n\n

Jaynes, E. T. (1957). Information theory and statistical mechanics. The Physical Review, 106(4),
620-630.
<\/p>\n\n\n\n

Jeziorski, J., Brandt, R., Evans, J. H., Campana, W., Kalichman, M., Thompson, E., Goldstein, L.,
<\/p>\n\n\n\n

Koch, C., & Muotri, A. R. (2023). Brain organoids, consciousness, ethics and moral status.
<\/p>\n\n\n\n

Jonas, H. (2001). The phenonomenon of life: Towards a philosophical biology. Northwestern
University Press.<\/p>\n\n\n\n

Kagan, B. J., Kitchen, A. C., Tran, N. T., Habibollahi, F., Khajehnejad, M., Parker, B. J., Bhat, A.,
Rollo, B., Razi, A., & Friston, K. J. (2022). In vitro neurons learn and exhibit sentience
when embodied in a simulated game-world. Neuron, 110(23), 3952-3969 e3958.
https:\/\/doi.org\/10.1016\/j.neuron.2022.09.001
<\/p>\n\n\n\n

Kagan, B. J., Razi, A., Bhat, A., Kitchen, A. C., Tran, N. T., Habibollahi, F., Khajehnejad, M., Parker,
B. J., Rollo, B., & Friston, K. J. (2023). Scientific communication and the semantics of
sentience. Neuron, 111(5), 606-607. https:\/\/doi.org\/10.1016\/j.neuron.2023.02.008
<\/p>\n\n\n\n

Kalra, A. P., Benny, A., Travis, S. M., Zizzi, E. A., Morales-Sanchez, A., Oblinsky, D. G., Craddock, T.
J. A., Hameroff, S. R., MacIver, M. B., Tuszynski, J. A., Petry, S., Penrose, R., & Scholes, G.
D. (2023). Electronic Energy Migration in Microtubules. ACS Cent Sci, 9(3), 352-361.
https:\/\/doi.org\/10.1021\/acscentsci.2c01114
<\/p>\n\n\n\n

Kirchhoff, M. D., & Froese, T. (2017). Where There Is Life There Is Mind: In Support of a Strong
Life-Mind Continuity Thesis. Entropy, 19(4). https:\/\/doi.org\/ARTN 169
10.3390\/e19040169
<\/p>\n\n\n\n

Kirkpatrick, K. L. (2022). Biological computation: hearts and flytraps. J Biol Phys, 48(1), 55-78.
https:\/\/doi.org\/10.1007\/s10867-021-09590-9
<\/p>\n\n\n\n

Kiverstein, J. (2020). Free energy and the self: an ecological-enactive interpretation. Topoi, 39,
559-574.
<\/p>\n\n\n\n

Kleene, S. C. (1956). Representation of events in nerve nets and finite automata. In C. Shannon &
J. McCarthy (Eds.), Automata Studies (pp. 3-42). Princeton University Press.
<\/p>\n\n\n\n

Kleiner, J. (2024). Consciousness requires mortal computation. https:\/\/arxiv.org\/abs\/2403.03925
<\/p>\n\n\n\n

Kleiner, J., & Ludwig, T. (2023). If consciousness is dynamically relevant, artificial intelligence isn\u2019t conscious. https:\/\/arxiv.org\/abs\/2304.05077
<\/p>\n\n\n\n

Koch, C., & Hepp, K. (2006). Quantum mechanics in the brain. Nature, 440(7084), 611.
https:\/\/doi.org\/10.1038\/440611a
<\/p>\n\n\n\n

Kuhn, R. L. (2024). A landscape of consciousness: Toward a taxonomy of explanations and implications. Progress in Biophysics and Molecular Biology, 190, 28-169.
<\/p>\n\n\n\n

Kurzweil, R. (2005). The singularity is near. Viking.
<\/p>\n\n\n\n

Kurzweil, R. (2024). The singularity is nearer: When we merge with AI. Bodley Head.
<\/p>\n\n\n\n

Lamme, V. A. (2010). How neuroscience will change our view on consciousness. Cognitive Neuroscience, 1(3), 204-240.
<\/p>\n\n\n\n

Lamme, V. A. (2018). Challenges for theories of consciousness: seeing or knowing, the missing ingredient and how to deal with panpsychism. Philosophical Transactions of the Royal Society B-Biological Sciences, 373, 20170344.
<\/p>\n\n\n\n

Landauer, R. (1961). Irreversibility and heat generation in the computing process. IBM Journal of Research and Development, 5(3), 183-191.
<\/p>\n\n\n\n

Lane, N. (2015). The vital question: Why is life the way it is? Profile Books.
<\/p>\n\n\n\n

Lane, N. (2022). Transformer: The deep chemistry of life and death. WW Norton \/ Profile.
<\/p>\n\n\n\n

Legg, S., & Hutter, M. (2007). A collection of definitions of intelligence. Frontiers in Artificial Intelligence and Applications, 157, 17-24. https:\/\/arxiv.org\/abs\/0706.3639v1
<\/p>\n\n\n\n

Levin, M. (2023). Darwin\u2019s agential materials: evolutionary implications of multiscale competency in developmental biology. Cellular and Molecular Life Sciences, 6.
https:\/\/doi.org\/https:\/\/doi.org\/10.1007\/s00018-023-04790-z
<\/p>\n\n\n\n

Levin, M. (2025). Ingressing Minds: Causal Patterns Beyond Genetics and Environment in Natural, Synthetic, and Hybrid Embodiments. PsyArXiv.
<\/p>\n\n\n\n

Li, N., Lu, D., Yang, L., Tao, H., Xu, Y., Wang, C., Fu, L., Liu, H., Chummum, Y., & Zhang, S. (2018). Nuclear Spin Attenuates the Anesthetic Potency of Xenon Isotopes in Mice: Implications for the Mechanisms of Anesthesia and Consciousness. Anesthesiology, 129(2), 271-277.
https:\/\/doi.org\/10.1097\/ALN.0000000000002226
<\/p>\n\n\n\n

Llinas, R. R., Ribary, U., Contreras, D., & Pedroarena, C. (1998). The neuronal basis for consciousness. Philos Trans R Soc Lond B Biol Sci, 353, 1841-1849.
<\/p>\n\n\n\n

Long, R., Sebo, J., Butlin, P., Finlinson, K., Fish, K., Harding, J., Pfau, J., Sims, T., Birch, J., & Chalmers, D. (2024). Taking AI welfare seriously. https:\/\/arxiv.org\/abs\/2411.00986
<\/p>\n\n\n\n

Mashour, G. A., Roelfsema, P., Changeux, J. P., & Dehaene, S. (2020). Conscious Processing and the Global Neuronal Workspace Hypothesis. Neuron, 105(5), 776-798. https:\/\/doi.org\/10.1016\/j.neuron.2020.01.026
<\/p>\n\n\n\n

Maturana, H., & Varela, F. (1980). Autopoiesis and Cognition: The Realization of the Living (Vol. 42). D. Reidel.
<\/p>\n\n\n\n

McCulloch, W. S., & Pitts, W. (1943). A logical calculus of the ideas immanent in nervous activity.
Bulletin of Mathematical Biology, 52, 99-115.
<\/p>\n\n\n\n

McFadden, J. (2020). Integrating information in the brain’s EM field: the cemi field theory of consciousness. Neurosci Conscious, 2020(1), niaa016. https:\/\/doi.org\/10.1093\/nc\/niaa016
<\/p>\n\n\n\n

McFadden, J., & Al-Khalili, J. (2018). The origins of quantum biology. Proc Math Phys Eng Sci, 474(2220), 20180674. https:\/\/doi.org\/10.1098\/rspa.2018.0674
<\/p>\n\n\n\n

McMillen, P., & Levin, M. (2024). Collective intelligence: A unifying concept for integrating
biology across scales and substrates. 7, 378. https:\/\/doi.org\/https:\/\/doi.org\/10.1038\/s42003-024-06037-4
<\/p>\n\n\n\n

Mead, C. (2020). How we created neuromorphic engineering. Nature Electronics, 3, 434-435.
<\/p>\n\n\n\n

Merolla, P. A., Arthur, J. V., Alvarez-Icaza, R., Cassidy, A. S., Sawada, J., Akopyan, F., Jackson, B. L., Imam, N., Guo, C., Nakamura, Y., Brezzo, B., Vo, I., Esser, S. K., Appuswamy, R., Taba, B., Amir, A., Flickner, M. D., Risk, W. P., Manohar, R., & Modha, D. S. (2014). Artificial brains. A million spiking-neuron integrated circuit with a scalable communication network and interface. Science, 345(6197), 668-673. https:\/\/doi.org\/10.1126\/science.1254642
<\/p>\n\n\n\n

Metzinger, T. (2021). Artificial suffering: An argument for a global moratorium on synthetic phenomenology. Journal of Artificial Intelligence and Consciousness, 8(1), 1024.
<\/p>\n\n\n\n

Metzinger, T. (2024). The elephant and the blind: The Experience of Pure Consciousness:
Philosophy, Science, and 500+ Experiential Reports. MIT Press.
<\/p>\n\n\n\n

Mitchell, K. J. (2023). Free agents: How evolution gave us free will. Princeton University Press.
<\/p>\n\n\n\n

Mitchell, M. (2024). The metaphors of artificial intelligence. Science, 386(6723), eadt6140.
https:\/\/doi.org\/10.1126\/science.adt6140
<\/p>\n\n\n\n

Mitchell, M., & Krakauer, D. C. (2023). The debate over understanding in AI\u2019s large language models. Proc Natl Acad Sci U S A, 120(13), e2215907120. https:\/\/doi.org\/https:\/\/doi.org\/10.1073\/pnas.2215907120
<\/p>\n\n\n\n

Mohseni, M., Omar, Y., & Engel, G. S. (Eds.). (2013). Quantum effects in biology. Cambridge University Press.<\/p>\n\n\n\n

Morales Pantoja, I. E., Smirnova, L., Muotri, A. R., Wahlin, K. J., Kahn, J., Boyd, J. L., Gracias, D. H., Harris, T. D., Cohen-Karni, T., Caffo, B. S., Szalay, A. S., Han, F., Zack, D. J., Etienne Cummings, R., Akwaboah, A., Romero, J. C., Alam El Din, D. M., Plotkin, J. D., Paulhamus, B. L., . . . Hartung, T. (2023). First Organoid Intelligence (OI) workshop to form an OI community. Front Artif Intell, 6, 1116870. https:\/\/doi.org\/10.3389\/frai.2023.1116870<\/p>\n\n\n\n

Mudrik, L., Boly, M., Dehaene, S., Fleming, S. M., Lamme, V., Seth, A. K., & Melloni, L. (2025). Unpacking the complexities of consciousness: Theories and reflections. Neurosci Biobehav Rev, 170, 106053.
https:\/\/doi.org\/https:\/\/doi.org\/10.1016\/j.neubiorev.2025.106053
<\/p>\n\n\n\n

Nagel, T. (1974). What is it like to be a bat? Philosophical Review, 83, 435-450.
<\/p>\n\n\n\n

Nave, K. (2025). A drive to survive: The free energy principle and the meaning of life. MIT Press.
<\/p>\n\n\n\n

Neven, H., Zalcman, A., Read, P. K., K., van der Molen, T., Bouwmeester, D., Bodnia, E., Turin, L.,
& Koch, C. (2024). Testing the conjecture that quantum processes create conscious
experience. Entropy, 26(6), 460.
<\/p>\n\n\n\n

O’Regan, J. K., & No\u00eb, A. (2001). A sensorimotor account of vision and visual consciousness. Behav Brain Sci, 24(5), 939-973; discussion 973-1031.
http:\/\/www.ncbi.nlm.nih.gov\/entrez\/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Citation&list_uids=12239892
<\/p>\n\n\n\n

Ororbia, A., & Friston, K., J. (2023). Mortal computation: a foundation for biomimetic intelligence. https:\/\/arxiv.org\/abs\/2311.09589
<\/p>\n\n\n\n

Parr, T., Pezzulo, G., & Friston, K., J. (2022). Active inference: the free energy principle in mind, brain, and behavior. MIT Press.
<\/p>\n\n\n\n

Penrose, R. (1989). The Emperor\u2019s New Mind. Oxford University Press.
<\/p>\n\n\n\n

Piccinini, G. (2018). Non-computational functionalism: Computation and the function of consciousness. Conference of Cognitive Computational Neuroscience,
<\/p>\n\n\n\n

Piccinini, G. (2020). Neurocognitive mechanisms. Oxford University Press.
<\/p>\n\n\n\n

Piccinini, G. (2023). The computational theory of mind. Cambridge University Press.
<\/p>\n\n\n\n

Polger, T., & Shapiro, L. (2016). The multiple realization book. Oxford University Press.
<\/p>\n\n\n\n

Prinz, J. (2012). The conscious brain: How attention engenders experience. Oxford University
Press.
<\/p>\n\n\n\n

Putnam, H. (1967). Psychological predicates. In W. H. Capitan & D. D. Merrill (Eds.), Art, Mind,
and Religion (pp. 37-48). Pittsburgh University Press.
<\/p>\n\n\n\n

Putnam, H. (1975). Mind, language, and reality. Cambridge University Press.
<\/p>\n\n\n\n

Putnam, H. (1988). Representation and reality. MIT Press.
<\/p>\n\n\n\n

Pylyshyn, Z. (1980). The \u2018causal power\u2019 of machines. Behavioral and Brain Sciences, 3, 442-444.
<\/p>\n\n\n\n

Rao, R. P., & Ballard, D. H. (1999). Predictive coding in the visual cortex: a functional
interpretation of some extra-classical receptive-field effects. Nat Neurosci, 2(1), 79-87.
https:\/\/doi.org\/10.1038\/4580
<\/p>\n\n\n\n

Rosas, F. E., Geiger, B. C., Luppi, A. I., Seth, A. K., Polani, D., Gastper, M., & Mediano, P. A. M.
(2024). Software in the natural world: A computational approach to hierarchical
emergence. https:\/\/arxiv.org\/abs\/2402.09090
<\/p>\n\n\n\n

Russell, S., & Norwig, P. (2003). Artificial Intelligence: A modern approach. Prentice Hall.
<\/p>\n\n\n\n

Schneider, S. (2019). Artificial You: AI and the Future of Your Mind. Princeton University Press.
<\/p>\n\n\n\n

Schr\u00f6dinger, E. (1944). What is life? Cambridge University Press.<\/p>\n\n\n\n

Schuman, C. D., Kulkarni, S. R., Parsa, M., Mitchell, J. P., Date, P., & Kay, B. (2022). Opportunities
for neuromorphic computing algorithms and applications. Nat Comput Sci, 2(1), 10-19.
https:\/\/doi.org\/10.1038\/s43588-021-00184-y
<\/p>\n\n\n\n

Schwitzgebel, E., & Garza, M. (2023). Designing AI with Rights, Consciousness, Self-Respect, and
Freedom. In F. Lara & J. Deckers (Eds.), Ethics of Artificial Intelligence. (pp. 459-479).
Springer Nature.
<\/p>\n\n\n\n

Searle, J. (1980). Minds, brains, and programs. Behavioral and Brain Sciences, 3(3), 417-457.
<\/p>\n\n\n\n

Searle, J. (1990). Is the brain a digital computer? Proceedings of the American Philosophical
Society, 64(3), 21-37.
<\/p>\n\n\n\n

Searle, J. (2007). Dualism revisited. Jourmal of Physiology – Paris, 101, 169-178.
<\/p>\n\n\n\n

Searle, J. (2017). Biological naturalism. In S. Schneider & M. Velmans (Eds.), The Blackwell
Companion to Consciousness. John Wiley and Sons Ltd.<\/p>\n\n\n\n

Sengupta, B., Stemmler, M. B., & Friston, K. J. (2013). Information and efficiency in the nervous
system–a synthesis. PLoS Comput Biol, 9(7), e1003157.
https:\/\/doi.org\/10.1371\/journal.pcbi.1003157
<\/p>\n\n\n\n

Seth, A. K. (2009). The strength of weak artificial consciousness. Journal of Machine
Consciousness, 1(1), 71-82.
<\/p>\n\n\n\n

Seth, A. K. (2012). Putting Descartes before the horse: Quantum theories of consciousness:
Comment on “Consciousness, biology, and quantum hypotheses” by Baars & Edelman
[Comment]. Physics of life reviews, 9(3), 297-298; discussion 306-297.
https:\/\/doi.org\/10.1016\/j.plrev.2012.07.005
<\/p>\n\n\n\n

Seth, A. K. (2013). Interoceptive inference, emotion, and the embodied self. Trends Cogn Sci,
17(11), 565-573. https:\/\/doi.org\/10.1016\/j.tics.2013.09.007
<\/p>\n\n\n\n

Seth, A. K. (2015). The cybernetic bayesian brain: from interoceptive inference to sensorimotor
contingencies. In J. M. Windt & T. Metzinger (Eds.), Open MIND (pp. 35(T)). MIND Group.
https:\/\/doi.org\/10.15502\/9783958570108
<\/p>\n\n\n\n

Seth, A. K. (2016). The real problem. Aeon
<\/p>\n\n\n\n

Seth, A. K. (2019). Being a beast machine: The origins of selfhood in control-oriented
interoceptive inference. In M. Colombo, L. Irvine, & M. Stapleton (Eds.), Andy Clark and
his Critics (pp. 238-254). Wiley-Blackwell.
<\/p>\n\n\n\n

Seth, A. K. (2021). Being You: A New Science of Consciousness. Faber & Faber.
<\/p>\n\n\n\n

Seth, A. K. (2023). Why conscious AI is a bad, bad idea. Nautilus.
<\/p>\n\n\n\n

Seth, A. K., & Bayne, T. (2022). Theories of consciousness. Nat Rev Neurosci, 23(7), 439-452.
https:\/\/doi.org\/10.1038\/s41583-022-00587-4
<\/p>\n\n\n\n

Seth, A. K., & Friston, K. J. (2016). Active interoceptive inference and the emotional brain. Philosophical Transactions of the Royal Society B-Biological Sciences, 371(1708).
https:\/\/doi.org\/ARTN 2016000710.1098\/rstb.2016.0007
<\/p>\n\n\n\n

Seth, A. K., & Tsakiris, M. (2018). Being a Beast Machine: The Somatic Basis of Selfhood. Trends Cogn Sci, 22(11), 969-981. https:\/\/doi.org\/10.1016\/j.tics.2018.08.008
<\/p>\n\n\n\n

Shagir, O. (2010). The rise and fall of computational functionalism. In Hilary Putnam (pp. 220 250). Cambridge University Press.
<\/p>\n\n\n\n

Shanahan, M. (2016). Conscious exotica. Aeon. https:\/\/aeon.co\/essays\/beyond-humans-what
other-kinds-of-minds-might-be-out-there<\/p>\n\n\n\n

Shanahan, M. (2024). Simulacra as conscious exotica. Inquiry. https:\/\/doi.org\/https:\/\/www.tandfonline.com\/doi\/full\/10.1080\/0020174X.2024.2434860
<\/p>\n\n\n\n

Shanahan, M., McDonell, K., & Reynolds, L. (2023). Role play with large language models. Nature, 623(7987), 493-498. https:\/\/doi.org\/10.1038\/s41586-023-06647-8
<\/p>\n\n\n\n

Shanahan, M., & Singler, B. (2024). Existential conversations with large language models: content, community, and culture. ArXiv, 2411.13223 https:\/\/doi.org\/https:\/\/doi.org\/10.48550\/arXiv.2411.13223
<\/p>\n\n\n\n

Shannon, C., & Weaver, W. (1949). The mathematical theory of communication. The University of Illinois Press.
<\/p>\n\n\n\n

Shapiro, L. (2008). How to test for multiple realization. Philosophy of Science, 75, 514-525.
<\/p>\n\n\n\n

Shiller, D. (2024). Functionalism, integrity, and digital consciousness. Synthese, 203, 47.
<\/p>\n\n\n\n

Shugg, W. (1968). The cartesian beast-machine in english literature (1663 – 1750). Journal of the History of ideas, 29(2), 279-292. http:\/\/www.jstor.org\/stable\/2708581
<\/p>\n\n\n\n

Simon, H. (1988). The Sciences of the Artificial (3rd ed.). MIT Press.
<\/p>\n\n\n\n

Simons, D. J., & Levin, D. T. (1997). Change blindness. Trends Cogn Sci, 1(7), 261-267.
https:\/\/doi.org\/10.1016\/S1364-6613(97)01080-2
<\/p>\n\n\n\n

Singer, P. (1975). Animal liberation: A new ethics for our treatment of animals. Harper Collins.
<\/p>\n\n\n\n

Solms, M. (2021). The Hidden Spring: A Journey to the Source of Consciousness. Profile Books.
<\/p>\n\n\n\n

Stade, E. C., Stirman, S. W., Ungar, L. H., Boland, C. L., Schwartz, H. A., Yaden, D. B., Sedoc, J.,
DeRubeis, R. J., Willer, R., & Eichstaedt, J. C. (2024). Large language models could change the future of behavioral healthcare: a proposal for responsible development and evaluation. Npj Ment Health Res, 3(1), 12. https:\/\/doi.org\/10.1038\/s44184-024-00056-z<\/p>\n\n\n\n


Sterling, P. (2012). Allostasis: a model of predictive regulation. Physiol Behav, 106(1), 5-15.
https:\/\/doi.org\/10.1016\/j.physbeh.2011.06.004<\/p>\n\n\n\n


Stewart, J., Gapenne, O., & Di Paolo, E. A. (2010). Enaction: Toward a new paradigm for cognitive science. MIT Press.<\/p>\n\n\n\n


Thompson, A. (1998). On the automatic design of robust electronics through artificial evolution.<\/p>\n\n\n\n


In M. Sipper, D. Mange, & A. Perez-Uribe (Eds.), Evolvable systems: From biology to
hardware (Vol. 1478, pp. 13-24). Springer.<\/p>\n\n\n\n


Thompson, E. (2007). Mind in life: Biology, phenomenology, and the sciences of mind. Harvard
University Press.<\/p>\n\n\n\n


Thompson, E. (2022). Could all life be sentient? Journal of Consciousness Studies, 29, 229-265.
<\/p>\n\n\n\n

Thompson, E., & Varela, F. J. (2001). Radical embodiment: neural dynamics and consciousness.
Trends Cogn Sci, 5(10), 418-425. http:\/\/www.ncbi.nlm.nih.gov\/entrez\/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Citation&list_uids=11707380
<\/p>\n\n\n\n

Tononi, G., Boly, M., Massimini, M., & Koch, C. (2016). Integrated information theory: from
consciousness to its physical substrate. Nat Rev Neurosci, 17(7), 450-461.
https:\/\/doi.org\/10.1038\/nrn.2016.44<\/p>\n\n\n\n

Tononi, G., & Edelman, G. M. (1998). Consciousness and complexity. Science, 282(5395), 1846-1851. http:\/\/www.ncbi.nlm.nih.gov\/entrez\/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Citati
on&list_uids=9836628<\/p>\n\n\n\n

    \n
  1. <\/li>\n<\/ol>\n\n\n\n

    Turing, A. M. (1936). On computable numbers, with an application to the Entscheidungsproblem. Proceedings of the London Mathematical Society, 42, 230-265.
    <\/p>\n\n\n\n

    Turing, A. M. (1950). Computing Machinery and Intelligence. Mind, 49, 433-460.
    <\/p>\n\n\n\n

    Turkle, S. (2021). The empathy diaries: a memoir. Penguin.
    <\/p>\n\n\n\n

    Ullman, B. (2022). Analog computing (2nd ed.). De Gruyter Oldenbourg.<\/p>\n\n\n\n

    Van Gelder, T. (1995). What might cognition be if not computation? Journal of Philosophy, 92(7), 345-381.<\/p>\n\n\n\n

    VanRullen, R., & Kanai, R. (2021). Deep learning and the Global Workspace Theory. Trends Neurosci. https:\/\/doi.org\/10.1016\/j.tins.2021.04.005
    <\/p>\n\n\n\n

    Varela, F. J. (1996). Neurophenomenology: A methodological remedy for the hard problem.
    Journal of Consciousness Studies, 3, 330-350.<\/p>\n\n\n\n


    Varela, F. J., Thompson, E., & Rosch, E. (1993). The embodied mind: Cognitive science and human
    experience. MIT Press.<\/p>\n\n\n\n


    V\u00e9liz, C. (2021). Privacy is power. Penguin.
    <\/p>\n\n\n\n

    Vong, W. K., Wang, W., Orhan, A. E., & Lake, B. M. (2024). Grounded language acquisition through the eyes and ears of a single child. Science, 383(6682), 504-511. https:\/\/doi.org\/10.1126\/science.adi1374
    <\/p>\n\n\n\n

    Wagenaar, W. A., & Sagaria, S. D. (1975). Misperception of exponential growth. Percept Psychophys, 18, 416-422.
    <\/p>\n\n\n\n

    Whyte, C. J., Corcoran, A. W., Robinson, J., Smith, R., Friston, K., J., Seth, A. K., & Hohwy, J. (2024). To see is to look: The minimal theory of consciousness implicit in active inference. https:\/\/arxiv.org\/abs\/2410.06633
    <\/p>\n\n\n\n

    Wiese, W. (2024). Artificial consciousness: a perspective from the free energy principle. Philosophical Studies. https:\/\/doi.org\/https:\/\/doi.org\/10.1007\/s11098-024-02182-y
    <\/p>\n\n\n\n

    Wimsatt, W. C. (1986). Developmental constraints, generative entrenchment, and the innate acquired distinction. In W. Bechtel (Ed.), Integrating Scientific Disciplines, Science and Philoso- phy (Vol. 2, pp. 185-208). Springer.<\/p>\n\n\n\n

    <\/p>\n","protected":false},"featured_media":5886,"template":"","meta":{"_acf_changed":false},"event_publishing_tags":[426,620,1076,84,397,276,524,415,280,1204,175,94,766,93,63,101,400,64,92,281,771,174,125,736,1171,1078,76,395,960,340,765,177,958,107,847,846,425,609,61,1170,338,96,176,396],"kategori":[305],"class_list":["post-5847","blog_content","type-blog_content","status-publish","has-post-thumbnail","hentry","event_publishing_tags-artificial","event_publishing_tags-artificial-consciousness","event_publishing_tags-artificial-general-intelligence","event_publishing_tags-artificial-intelligence","event_publishing_tags-autopoiesis","event_publishing_tags-beden","event_publishing_tags-bedenlenme","event_publishing_tags-bedenlenmis-bilis","event_publishing_tags-bilgisayar","event_publishing_tags-bilgisayar-meatforu","event_publishing_tags-bilinc","event_publishing_tags-bilis","event_publishing_tags-bilis-felsefesi","event_publishing_tags-bilissel-bilim","event_publishing_tags-biology","event_publishing_tags-biyoloji","event_publishing_tags-body","event_publishing_tags-cognition","event_publishing_tags-cognitive-science","event_publishing_tags-computer","event_publishing_tags-computer-metaphor","event_publishing_tags-consciousness","event_publishing_tags-embodied-cognition","event_publishing_tags-embodiment","event_publishing_tags-ethics-of-articifical-intelligence","event_publishing_tags-genel-yapay-zeka","event_publishing_tags-mind","event_publishing_tags-mind-body-problem","event_publishing_tags-ongormeli-isleme","event_publishing_tags-philosophy-of-artificial-intelligence","event_publishing_tags-philosophy-of-cognition","event_publishing_tags-philosophy-of-mind","event_publishing_tags-predictive-processing","event_publishing_tags-robot","event_publishing_tags-robot-bilisi","event_publishing_tags-robot-cognition","event_publishing_tags-yapay","event_publishing_tags-yapay-bilinc","event_publishing_tags-yapay-zeka","event_publishing_tags-yapay-zeka-etigi","event_publishing_tags-yapay-zeka-felsefesi","event_publishing_tags-zihin","event_publishing_tags-zihin-felsefesi","event_publishing_tags-zihin-beden-problemi","kategori-ceviri"],"acf":[],"_links":{"self":[{"href":"https:\/\/cog-ist.com\/en\/wp-json\/wp\/v2\/blog_content\/5847","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/cog-ist.com\/en\/wp-json\/wp\/v2\/blog_content"}],"about":[{"href":"https:\/\/cog-ist.com\/en\/wp-json\/wp\/v2\/types\/blog_content"}],"version-history":[{"count":35,"href":"https:\/\/cog-ist.com\/en\/wp-json\/wp\/v2\/blog_content\/5847\/revisions"}],"predecessor-version":[{"id":5889,"href":"https:\/\/cog-ist.com\/en\/wp-json\/wp\/v2\/blog_content\/5847\/revisions\/5889"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/cog-ist.com\/en\/wp-json\/wp\/v2\/media\/5886"}],"wp:attachment":[{"href":"https:\/\/cog-ist.com\/en\/wp-json\/wp\/v2\/media?parent=5847"}],"wp:term":[{"taxonomy":"event_publishing_tags","embeddable":true,"href":"https:\/\/cog-ist.com\/en\/wp-json\/wp\/v2\/event_publishing_tags?post=5847"},{"taxonomy":"kategori","embeddable":true,"href":"https:\/\/cog-ist.com\/en\/wp-json\/wp\/v2\/kategori?post=5847"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}