Öz
Bu makale, teknik modelleri (sembolik, bağlantısal, hibrit) Wittgenstein’ın dil felsefesi ile ilişkilendirerek yapay zekadaki (YZ) temsil sorununu incelemeyi ve bunların bilişsel bilim ve psikiyatri pratiği için çıkarımlarını vurgulamayı amaçlamaktadır. YZ’de temsil sorunu, güvenilir ve yorumlanabilir sistemler geliştirmede karşılaşılan temel zorluklardan biridir. Bu sorun, YZ’nin tanımıyla yakından ilişkilidir ve rasyonel davranış, insan benzeri düşünme ve insan benzeri davranış yaklaşımları üzerinden ele alınabilir. Sembolik, bağlantısal ve hibrit modeller gibi temsil yöntemleri, bilginin kodlanması, işlenmesi ve yorumlanması için farklı mekanizmalar sunar. İstatistiksel yaklaşımlardan derin öğrenme ve transformer tabanlı mimarilere geçiş, büyük dil modellerinin doğal dili işleme yeteneklerinde önemli ilerlemeler sağlamıştır. Wittgenstein’ın “resim kuramı” ve “dil oyunları” yaklaşımları, bu modellerde anlam inşasını anlamak için kavramsal bir çerçeve sunar. Açıklanabilir YZ, şeffaflık ve güvenin artırılması açısından kritik bir yaklaşım olarak öne çıkmaktadır. Bilişsel bilim perspektifinden, YZ’nin insan bilişsel süreçlerini ne ölçüde taklit edebileceği konusu önemli tartışmalardan biridir. Çoklu görev öğrenme, az örnekle öğrenme ve bağlamsal anlam çıkarma gibi yetenekler, insan zihninin öğrenme biçimleriyle benzerlik gösterse de, YZ hala kültürel bağlamı, yaşam formunu ve çoklu modalite üzerinden anlamı tam olarak yakalamakta sınırlıdır. Gelecekte kuantum bilişim, sinirbilim verilerinin (örn, elektroensefalogram) entegrasyonu ve çoklu modalite sistemleri, temsil stratejilerini yeniden şekillendirme ve YZ’yi bilişsel bilim ile bütünleştirerek daha insan benzeri, bağlamı yakalayabilen yetenekler kazandırma potansiyeline sahiptir. Bununla birlikte, psikiyatri uygulamalarında kullanılan sembolik, bağlantısal ve hibrit yaklaşımlar oldukça önemlidir. YZ’nin depresyon tespiti, şizofreni riskinin öngörülmesi, demans bakımında sosyal etkileşimin artırılması gibi çok çeşitli klinik uygulamaları mevcuttur. Psikoterapi ve dil temsili alanındaki gelişmelerin gelecekte YZ’nin psikiyatrideki işlevselliğini artıracağı öngörülmektedir. Genel olarak, temsil sorununa ilişkin felsefi, teknik ve klinik bakış açılarının bütünleştirilmesi, psikiyatride daha yorumlanabilir, daha güvenli ve klinik açıdan daha yararlı YZ sistemlerine giden bir yol haritası sunar.
Anahtar Kelimeler:
Bilişsel bilim, büyük dil modelleri, psikiyatri, temsil, Wittgenstein, yapay zeka
Kaynaklar
1Turing AM. Computing machinery and intelligence. Mind. 1950;59(236):433-460.
2Brown TB, Mann B, Ryder N, Subbiah M, Kaplan J, Dhariwal P, et al. Language models are few-shot learners. Adv Neural Inf Process Syst. 2020;33:1877-1901.
3Bender EM, Koller A. Climbing towards NLU: on meaning, form, and understanding in the age of data. In: Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics; 2020:5185-5198.
4Devlin J, Chang MW, Lee K, Toutanova K. BERT: pre-training of deep bidirectional transformers for language understanding. In: Proceedings of NAACL-HLT. 2019;2019:4171-4186.
5Lake BM, Baroni M. Generalization without systematicity: on the compositional skills of sequence-to-sequence recurrent networks. In: Proceedings of the 35th International Conference on Machine Learning (ICML 2018). 2018:2879-2888.
6Schneider S, Leyer M. Me or information technology? Adoption of artificial intelligence in the delegation of personal strategic decisions. Manag Decis Econ. 2019;40:223-231.
7Topol EJ. High-performance medicine: the convergence of human and artificial intelligence. Nat Med. 2019;25(1):44-56.
8World Health Organization. Mental Health Atlas 2020 [Internet]. 2021 [cited 2025 Aug 19]. Available from: https://www.who.int/publications/i/item/9789240036703
9Russell SJ, Norvig P. artificial intelligence: a modern approach. 4th ed. Hoboken, NJ: Pearson; 2021. Available link: http://lib.ysu.am/disciplines_bk/efdd4d1d4c2087fe1cbe03d9ced67f34.pdf
10Nilsson NJ. Artificial intelligence: a new synthesis. San Francisco, CA: Morgan Kaufmann; 1998.
11Simon HA, Newell A. Human problem solving. Englewood Cliffs, NJ: Prentice-Hall; 1971. Available link: https://iiif.library.cmu.edu/file/Newell_box00049_fld04211_doc0001/Newell_box00049_fld04211_doc0001.pdf
12Shieber SM. The turing test: verbal behavior as the Hallmark of intelligence. Cambridge, MA: MIT Press; 2005;31(3).
13Brachman RJ, Levesque HJ. Knowledge representation and reasoning. San Francisco, CA: Morgan Kaufmann; 2004. Available link: https://bista.sites.dmi.unipg.it/didattica/CSP/knowledge-representation-and-reasoning-the-morgan-kaufmann-series-in-artificial-intelligence.pdf
14Wittgenstein L. Philosophical investigations. Oxford, UK: Basil Blackwell; 1953. Available link: https://static1.squarespace.com/static/54889e73e4b0a2c1f9891289/t/564b61a4e4b04eca59c4d232/1447780772744/Ludwig.Wittgenstein.-.Philosophical.Investigations.pdf
15McCarthy J. Circumscription—a form of non-monotonic reasoning. Artif Intell. 1980;13(1-2):27-39.
16Davis R, Shrobe H, Szolovits P. What is a knowledge representation? AI Mag. 1993;14(1):17-33.
17Hebb DO. The organization of behavior: a neuropsychological theory. New York, NY: John Wiley & Sons; 1949. Available link: https://pure.mpg.de/rest/items/item_2346268_3/component/file_2346267/content
18Rosenblatt F. The perceptron: a probabilistic model for information storage and organization in the brain. Psychol Rev. 1958;65(6):386-408.
19Rumelhart DE, McClelland JL, eds. Parallel distributed processing: explorations in the microstructure of cognition. Vol 1. Cambridge, MA: MIT Press; 1986. Available link: https://gwern.net/doc/ai/nn/1986-rumelhart-pdp-v1.pdf
20Goodfellow I, Bengio Y, Courville A. Deep learning. Cambridge, MA: MIT Press; 2016. Available link: https://repo.darmajaya.ac.id/4189/1/Deep%20learning_%20adaptive%20computation%20and%20machine%20learning%20%28%20PDFDrive%20%29.pdf
21Newell A. Unified theories of cognition. Cambridge, MA: Harvard University Press; 1990. Available link: https://iiif.library.cmu.edu/file/Newell_box00022_fld01532_doc0001/Newell_box00022_fld01532_doc0001.pdf
22Anderson JR. ACT: a simple theory of complex cognition. Am Psychol. 1996;51(4):355-365.
23Vaswani A, Shazeer N, Parmar N, Uszkoreit J, Jones L, Gomez AN, et al. Attention is all you need. Adv Neural Inf Process Syst. 2017;30:5998-6008.
24Jurafsky D, Martin JH. Speech and language processing [Internet]. 3rd ed. draft. 2025 [cited 2025 Aug 19]. Available from: https://web.stanford.edu/~jurafsky/slp3/
25Hochreiter S, Schmidhuber J. Long short-term memory. Neural Comput. 1997;9(8):1735-1780.
26Radford A, Wu J, Child R, Luan D, Amodei D, Sutskever I. Language models are unsupervised multitask learners. OpenAI Blog [Internet]. 2019 [cited 2025 Aug 19]. Available from: https://cdn.openai.com/better-language- models/language_models_are_unsupervised_multitask_learners.pdf
27Wittgenstein L. Tractatus logico-philosophicus. Pears DF, McGuinness BF (trans). London, UK: Routledge; 2001. Originally published 1921.
28Montague R. The proper treatment of quantification in ordinary English. In: Hintikka J, Moravcsik J, Suppes P (eds). Approaches to Natural Language. Dordrecht, NL: Reidel; 1973:221-242.
29Frege G. Über sinn und bedeutung. Z Philos Philos Krit. 1892;100(1):25-50.
30Besold TR, Garcez AA, Bader S, Bowman H, Domingos P, Hitzler P, et al. Neural- symbolic learning and reasoning: a survey and interpretation. arXiv [Internet]. 2017 [cited 2025 Aug 18].
31Zambak A, Vergauwen R. Artificial intelligence and agentive cognition: a logico- linguistic approach. Logique Analyse. 2009;52(205):57-96.
32Barrett DG, Morcos AS, Macke JH. Analyzing biological and artificial neural networks: challenges with opportunities for synergy? Curr Opin Neurobiol. 2019;55:55-64.
33Flesch T, Balaguer J, Dekker R, Nili H, Summerfield C. Comparing continual task learning in minds and machines. Proc Natl Acad Sci U S A. 2018;115(44):E10313-E10322.
34Srivastava N, Hinton G, Krizhevsky A, Sutskever I, Salakhutdinov R. Dropout: a simple way to prevent neural networks from overfitting. J Mach Learn Res. 2014;15:1929-1958.
35Hewitt J, Manning CD. A structural probe for finding syntax in word representations. In: Proceedings of NAACL-HLT 2019; 2019 Jun; Minneapolis, MN. Stroudsburg, PA: Association for Computational Linguistics; 2019:4129-4138.
36Ravichander A, Belinkov Y, Hovy E. Probing the probing paradigm: does probing accuracy entail task relevance? arXiv [Internet]. 2020 [cited 2025 Aug 19].
37Kaplan A, Haenlein M. Siri, Siri, in my hand: who’s the fairest in the land? Bus Horiz. 2019;62(1):15-25.
38Gunning D, Aha D. DARPA’s explainable artificial intelligence (XAI) program. AI Mag. 2019;40(2):44-58.
39Jobin A, Ienca M, Vayena E. The global landscape of AI ethics guidelines. Nat Mach Intell. 2019;1(9):389-399.
40Dellermann D, Ebel P, Söllner M, Leimeister JM. Hybrid intelligence. Bus Inf Syst Eng. 2019;61(5):637-643.
41Hutchins E. Cognition in the wild. Cambridge, MA: MIT Press; 1995.
42Miller GA. The cognitive revolution: a historical perspective. Trends Cogn Sci. 2003;7(3):141-144.
43Kiela D, Bulat L, Vero AL, Clark S. Virtual embodiment: a scalable long-term strategy for artificial intelligence research. arXiv [Internet]. 2016 [cited 2025 Aug 19].
44Inkster B, Sarda S, Subramanian V. An empathy-driven conversational AI agent (Wysa) for digital mental well-being. JMIR Mhealth Uhealth. 2018;6(11):e12106.
45Fitzpatrick KK, Darcy A, Vierhile M. Delivering CBT using a fully automated conversational agent (Woebot). JMIR Ment Health. 2017;4(2):e19.
46Vieira S, Pinaya WHL, Mechelli A. Using deep learning to investigate neuroimaging correlates of psychiatric disorders. Neurosci Biobehav Rev. 2017;74(Pt A):58-75.
47Bedi G, Carrillo F, Cecchi GA, Slezak DF, Sigman M, de la Iglesia H, et al. Automated analysis of free speech predicts psychosis onset. NPJ Schizophr. 2015;1:15030.
48Garety P, Ward T, Emsley R, Greenwood K, Freeman D, Fowler D, et al. Effects of slowMo on paranoia. JAMA Psychiatry. 2021;78(7):714-725.
49Moyle W, Jones CJ, Murfield JE, Thalib L, Beattie ERA, Shum DKH, et al. Robotic seal therapy in dementia. J Am Med Dir Assoc. 2017;18(9):766-773.
50Kargar BAH, Mahoor MH. eBear socially assistive robot pilot study. In: 2017 IEEE- RAS Int Conf Humanoid Robots; 2017.
51Floridi L, Cowls J, Beltrametti M, Chatila R, Chazerand P, Dignum V, et al. AI4People ethical framework. Minds Mach. 2018;28(4):689-707.
52Hariman K, Patel J, Whiting P. Artificial intelligence in psychiatry: are we ready yet? BJPsych Adv. 2019;25(5):279-286.
53Pham KT, Nabizadeh A, Selek S. Artificial intelligence and chatbots in psychiatry. Psychiatr Q. 2022;93(1):249-253.
54Ray A, Bhardwaj A, Malik YK, Singh S, Gupta R. Artificial intelligence and psychiatry: an overview. Asian J Psychiatr. 2022;70:103021.
55Gunning D. Explainable artificial intelligence (XAI). DARPA [Internet]. 2017 [cited 2025 Aug 18]. Available from: https://www.darpa.mil/program/explainable-artificial- intelligence
56Doshi-Velez F, Kim B. Towards a rigorous science of interpretable machine learning. arXiv [Internet]. 2017 [cited 2025 Aug 18]. Available from: https://arxiv.org/abs/1702.08608
57Biamonte J, Wittek P, Pancotti N, Rebentrost P, Wiebe N, Lloyd S. Quantum machine learning. Nature. 2017;549(7671):195-202.
58Kiperwasser E, Goldberg Y. Dependency parsing using bidirectional LSTM features. arXiv [Internet]. 2016 [cited 2025 Aug 19].
59Schuler KK. VerbNet: A broad-coverage, comprehensive verb lexicon [PhD dissertation]. Philadelphia (PA): University of Pennsylvania; 2005.
60Luck SJ. An introduction to the event-related potential technique. 2nd ed. Cambridge, MA: MIT Press; 2014.
61Kutas M, Federmeier KD. Thirty years of the N400. Annu Rev Psychol. 2011;62:621-647.
62Barsalou LW. Grounded cognition. Annu Rev Psychol. 2008;59:617-645.
63Belli H, Laçin S. Artificial intelligence in psychiatry: applications and challenges. Atlas J Med. 2025;5(14):224-233.