Název: Irelevantnosť Turingovho testu v súčasnom hlbokom učení
Variantní název:
- The irrelevance of the Turing test in current deep learning
Zdrojový dokument: Pro-Fil. 2021, roč. 22, č. 2, s. 28-44
Rozsah
28-44
-
ISSN1212-9097 (online)
Trvalý odkaz (DOI): https://doi.org/10.5817/pf21-2-2396
Trvalý odkaz (handle): https://hdl.handle.net/11222.digilib/144823
Type: Článek
Jazyk
Licence: CC BY-NC-ND 4.0 International
Upozornění: Tyto citace jsou generovány automaticky. Nemusí být zcela správně podle citačních pravidel.
Abstrakt(y)
Úlohou umelej inteligencie (UI) v Turingovom teste je imitovať človeka do takej miery, aby vyšetrovateľ nebol schopný rozlíšiť stroj od človeka. S príchodom hlbokého učenia (DL) (podkategória UI) sa však situácia mení, pretože sa tieto systémy namiesto simulovania ľudskej inteligencie zameriavajú na riešenie konkrétnych problémov. Z dôvodu, že tieto umelé systémy nesimulujú ľudskú inteligenciu, sa otvára otázka, či nie je Turingov test v problematike hlbokého učenia irelevantný. Na problém sa je možné pozrieť v troch častiach. Po prvé, sa je potrebné zamerať na aplikačné využitie Turingovho testu v Loebnerovej cene, v ktorej sú kladené otázky zamerané na aspekty ľudskej inteligencie – učenie, usudzovanie a porozumenie. Po druhé, je možné považovať za problém, že sa v Turingovom teste rozumie pod inteligenciou iba všeobecná ľudská inteligencia. Keďže ani DL touto formou inteligencie nedisponuje, je možné bez pochýb označiť túto UI za neinteligentnú? Nakoniec je otázne, či by vlastne malo zmysel, aby účelovo zameraná UI, akou je DL, absolvovala Turingov test, nakoľko samotný test žiadne ďalšie poznatky o analýze problémov alebo inteligencii neprináša.
The role of artificial intelligence in the Turing test is to imitate human beings to such an extent that people will not realize it is a machine. With the rise of deep learning (a subcategory of AI), the situation is changing rapidly as the new systems do not focus on imitating human intelligence but emphasize thorough solutions to specific issues. The main difference between predefined AI and deep learning (DL) is that these systems are self-learning and have verifiable results. Firstly, we need to analyse the application of the Turing test in the Loebner Prize because, there, the primary emphasis is on aspects of human intelligence – learning, reasoning and understanding. Secondly, in the Turing test, only general intelligence is considered, and this can be questionable. If DL does not possess this form of intelligence, by this reasoning, we should consider it unintelligent. However, is such understanding correct? The third and last aspect questions whether the Turing test is beneficial for an AI designed for specific tasks because the results do not bring any new data and conclusions.
Note
Tento príspevok vznikol vďaka podpore APVV-17-0064 - Analýza multidimenzionálnej podoby trans- a posthumanizmu.
Reference
[1] Bengio, Y. (2013): Representation Learning: A Review and New Perspectives, IEEE Transactions on Pattern Analysis and Machine Intelligence 35(8), 1–30, dostupné z: < 10.1109/TPAMI.2013.50 >.
[2] Fitch, W. T. (2014): Toward a computational framework for cognitive biology: Unifying approaches from cognitive neuroscience and comparative cognition, Physics of Life Reviews 11(3), 329–364, dostupné z: < https://doi-org.ezproxy.muni.cz/10.1016/j.plrev.2014.04.005 >. | DOI 10.1016/j.plrev.2014.04.005
[3] Flach, P. A. – Kakas, A. C. (2000): Abductive and Inductive Reasoning: Background and Issues, in Flach, P. A. – Kakas, A. C. (eds.) Abduction and Induction. Applied Logic Series 18, 1–27, dostupné z: < https://doi-org.ezproxy.muni.cz/10.1007/978-94-017-0606-3_1 >.
[4] Ford, K. M. – Hayes, P. J. (1998): On Computational Wings: Rethinking the Goals of Artificial Intelligence, Scientific American Presents 9, 78–83.
[5] French, R. M. (1990): Subcognition and the Limits of the Turing Test, Mind 99(393), 53–65, dostupné z: < https://doi-org.ezproxy.muni.cz/10.1093/mind/XCIX.393.53 >.
[6] French, R. M. (2000): The Turing Test: The First 50 Years, Trends in Cognitive Sciences 4(3), 115–122, dostupné z: < https://doi-org.ezproxy.muni.cz/10.1016/S1364-6613(00)01453-4 >.
[7] Gomes, L. (2014): Machine-Learning Maestro Michael Jordan on the Delusions of Big Data and Other Huge Engineering Efforts, IEEE Spectrum [online], 2014-10-20, [cit. 2021-03-31], dostupné z: < https://spectrum-ieee-org.ezproxy.muni.cz/machinelearning-maestro-michael-jordan-on-the-de-lusions-of-big-data-and-other-huge-engineering-efforts >.
[8] Goulden, R. – Nation, P. – Read, J. (1990): How Large Can a Receptive Vocabulary Be? Applied Linguistics 11(4), 341–363, dostupné z: < https://doi-org.ezproxy.muni.cz/10.1093/applin/11.4.341 >.
[9] Hriadel, O. (2020): Kognitívna (ne)uzavretosť umelej inteligencie (Deep Learning), in Cintula, I. – Mydlová K. – Kalivodová, V. (eds.) LOQUERE: interdisciplinárna doktorandská konferencia: zborník príspevkov, Univerzita sv. Cyrila a Metoda v Trnave, 105–117.
[10] Hsu, F-H. (2004): Behind Deep Blue: Building the Computer that Defeated the World Chess Champion, Princeton University Press.
[11] Chollet, F. (2019): Deep learning v jazyku Python, Grada.
[12] Kakas, A. – Michael, L. (2020): Abduction and Argumentation for Explainable Machine Learning: A Position Survey, Cornell University arXiv.org > cs > arXiv:2010.12896, 1–24, dostupné z: < https://arxiv.org/abs/2010.12896 >.
[13] Kruger, N. et al. (2013): Deep Hierarchies in the Primate Visual Cortex: What Can We Learn for Computer Vision?, IEEE Transactions on Pattern Analysis and Machine Intelligence 35(8), 1847–1871, dostupné z: < 10.1109/TPAMI.2012.272 >. | DOI 10.1109/TPAMI.2012.272
[14] Lacker, K. (2020): Giving GPT-3 a Turing Test, Kevin Lacker's blog [online], 2020-07-06, [cit. 2021-03-31], dostupné z: < https://lacker.io/ai/2020/07/06/giving-gpt-3-a-turing-test.html >.
[15] Levesque , H. J. (2017): Common Sense, the Turing Test, and the Quest for Real AI, The MIT Press.
[16] López-Rubio, E. (2018): Computational Functionalism for the Deep Learning Era, Minds and machines 28, 667–688, dostupné z: < https://doi-org.ezproxy.muni.cz/10.1007/s11023-018-9480-7 >.
[17] Mooney R. J. (2000): Integrating Abduction and Induction in Machine Learning, in Flach, P. A. – Kakas, A. C. (eds.) Abduction and Induction. Applied Logic Series 18, 181–191, dostupné z: < https://doi-org.ezproxy.muni.cz/10.1007/978-94-017-0606-3_12 >.
[18] Moor, J. H. (2001): The Status and Future of the Turing Test, Minds and Machines 11, 77–93, dostupné z: < https://doi-org.ezproxy.muni.cz/10.1023/A:1011218925467 >.
[19] Nguifo, E. M. (1998): Abduction and Induction in Learning Task: which needs the other? ACADEMIA – Accelerating the world's research 1–3, dostupné z: < https://scholar.google.com/scholar?hl=sk&as_sdt=0%2C5&q=Abduction+and+In-duction+in+Learning+Task%3A+which+needs+the+other+%3F&btnG= >.
[20] Otterlo, M. V. (2013): A Machine Learning View on Profiling, Privacy, Due Process and the Computational Turn The Philosophy of Law Meets the Philosophy of Technology, 41–65, dostupné z: < https://doi-org.ezproxy.muni.cz/10.4324/9780203427644 >.
[21] Parnas, D. L. (2017): The real risks of artificial intelligence, Communications of the ACM 60(10), 27–31, dostupné z: < https://doi-org.ezproxy.muni.cz/10.1145/3132724 >. | DOI 10.1145/3132724
[22] Searle, J. (1984): Mysl, mozek a věda, Mladá fronta, 1994.
[23] Schubbach, A. (2019): Judging machines: philosophical aspects of deep learning, Synthese 198, 1807–1827, dostupné z: < https://doi-org.ezproxy.muni.cz/10.1007/s11229-019-02167-z >.
[24] Silver, D. – Huang, A. – Maddison, C. et al. (2016): Mastering the game of Go with deep neural networks and tree search, Nature 529, 484–489, dostupné z: < https:doi-org.ezproxy.muni.cz/10.1038/nature16961 >.
[25] Slonim, N. – Bilu, Y. – Alzate, C. et al. (2021): An autonomous debating system, Nature 591, 379–384, dostupné z: < https:doi-org.ezproxy.muni.cz/10.1038/s41586-021-03215-w >.
[26] Tan C. et al. (2018): A Survey on Deep Transfer Learning, in Kůrková V. et al. (eds.) Artificial Neural Networks and Machine Learning – ICANN 2018. Lecture Notes in Computer Science 11141, 270–279, dostupné z: < https://doi-org.ezproxy.muni.cz/10.1007/978-3-030-01424-7_27 >.
[27] Toledo, A. et al. (2019): Automatic Argument Quality Assessment – New Datasets and Methods, Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP), D19-1564, 5625–5635, dostupné z: < https://doi-org.ezproxy.muni.cz/10.18653/v1/D19-1564 >.
[28] Turing, A. (1950): Počítacie stroje a inteligencia, Medzinárodná účastnícka spoločnosť Bradlo, 1992.
[29] Turing, A. (1950): Computing Machinery and Intelligence. Mind 59(236), 433–460, dostupné z: < https://doi-org.ezproxy.muni.cz/10.1093/mind/LIX.236.433 >.
[30] Vinyals, O. et al. (2016): Matching Networks for One Shot Learning, Advances in Neural Information Processing Systems 29, 1–12, dostupné z: < https://arxiv.org/abs/1606.04080 >.
[31] Yamins, D. – DiCarlo, J. (2016): Using goal-driven deep learning models to understand sensory cortex, Nat Neurosci 19, 356–365, dostupné z: < https://doi-org.ezproxy.muni.cz/10.1038/nn.4244 > | DOI 10.1038/nn.4244
[32] Zhou, Z-H. (2019): Abductive learning: towards bridging machine learning and logical reasoning, Sci. China Inf. Sci. 62(76101), dostupné z: < https://doi-org.ezproxy.muni.cz/10.1007/s11432-018-9801-4 >.
[2] Fitch, W. T. (2014): Toward a computational framework for cognitive biology: Unifying approaches from cognitive neuroscience and comparative cognition, Physics of Life Reviews 11(3), 329–364, dostupné z: < https://doi-org.ezproxy.muni.cz/10.1016/j.plrev.2014.04.005 >. | DOI 10.1016/j.plrev.2014.04.005
[3] Flach, P. A. – Kakas, A. C. (2000): Abductive and Inductive Reasoning: Background and Issues, in Flach, P. A. – Kakas, A. C. (eds.) Abduction and Induction. Applied Logic Series 18, 1–27, dostupné z: < https://doi-org.ezproxy.muni.cz/10.1007/978-94-017-0606-3_1 >.
[4] Ford, K. M. – Hayes, P. J. (1998): On Computational Wings: Rethinking the Goals of Artificial Intelligence, Scientific American Presents 9, 78–83.
[5] French, R. M. (1990): Subcognition and the Limits of the Turing Test, Mind 99(393), 53–65, dostupné z: < https://doi-org.ezproxy.muni.cz/10.1093/mind/XCIX.393.53 >.
[6] French, R. M. (2000): The Turing Test: The First 50 Years, Trends in Cognitive Sciences 4(3), 115–122, dostupné z: < https://doi-org.ezproxy.muni.cz/10.1016/S1364-6613(00)01453-4 >.
[7] Gomes, L. (2014): Machine-Learning Maestro Michael Jordan on the Delusions of Big Data and Other Huge Engineering Efforts, IEEE Spectrum [online], 2014-10-20, [cit. 2021-03-31], dostupné z: < https://spectrum-ieee-org.ezproxy.muni.cz/machinelearning-maestro-michael-jordan-on-the-de-lusions-of-big-data-and-other-huge-engineering-efforts >.
[8] Goulden, R. – Nation, P. – Read, J. (1990): How Large Can a Receptive Vocabulary Be? Applied Linguistics 11(4), 341–363, dostupné z: < https://doi-org.ezproxy.muni.cz/10.1093/applin/11.4.341 >.
[9] Hriadel, O. (2020): Kognitívna (ne)uzavretosť umelej inteligencie (Deep Learning), in Cintula, I. – Mydlová K. – Kalivodová, V. (eds.) LOQUERE: interdisciplinárna doktorandská konferencia: zborník príspevkov, Univerzita sv. Cyrila a Metoda v Trnave, 105–117.
[10] Hsu, F-H. (2004): Behind Deep Blue: Building the Computer that Defeated the World Chess Champion, Princeton University Press.
[11] Chollet, F. (2019): Deep learning v jazyku Python, Grada.
[12] Kakas, A. – Michael, L. (2020): Abduction and Argumentation for Explainable Machine Learning: A Position Survey, Cornell University arXiv.org > cs > arXiv:2010.12896, 1–24, dostupné z: < https://arxiv.org/abs/2010.12896 >.
[13] Kruger, N. et al. (2013): Deep Hierarchies in the Primate Visual Cortex: What Can We Learn for Computer Vision?, IEEE Transactions on Pattern Analysis and Machine Intelligence 35(8), 1847–1871, dostupné z: < 10.1109/TPAMI.2012.272 >. | DOI 10.1109/TPAMI.2012.272
[14] Lacker, K. (2020): Giving GPT-3 a Turing Test, Kevin Lacker's blog [online], 2020-07-06, [cit. 2021-03-31], dostupné z: < https://lacker.io/ai/2020/07/06/giving-gpt-3-a-turing-test.html >.
[15] Levesque , H. J. (2017): Common Sense, the Turing Test, and the Quest for Real AI, The MIT Press.
[16] López-Rubio, E. (2018): Computational Functionalism for the Deep Learning Era, Minds and machines 28, 667–688, dostupné z: < https://doi-org.ezproxy.muni.cz/10.1007/s11023-018-9480-7 >.
[17] Mooney R. J. (2000): Integrating Abduction and Induction in Machine Learning, in Flach, P. A. – Kakas, A. C. (eds.) Abduction and Induction. Applied Logic Series 18, 181–191, dostupné z: < https://doi-org.ezproxy.muni.cz/10.1007/978-94-017-0606-3_12 >.
[18] Moor, J. H. (2001): The Status and Future of the Turing Test, Minds and Machines 11, 77–93, dostupné z: < https://doi-org.ezproxy.muni.cz/10.1023/A:1011218925467 >.
[19] Nguifo, E. M. (1998): Abduction and Induction in Learning Task: which needs the other? ACADEMIA – Accelerating the world's research 1–3, dostupné z: < https://scholar.google.com/scholar?hl=sk&as_sdt=0%2C5&q=Abduction+and+In-duction+in+Learning+Task%3A+which+needs+the+other+%3F&btnG= >.
[20] Otterlo, M. V. (2013): A Machine Learning View on Profiling, Privacy, Due Process and the Computational Turn The Philosophy of Law Meets the Philosophy of Technology, 41–65, dostupné z: < https://doi-org.ezproxy.muni.cz/10.4324/9780203427644 >.
[21] Parnas, D. L. (2017): The real risks of artificial intelligence, Communications of the ACM 60(10), 27–31, dostupné z: < https://doi-org.ezproxy.muni.cz/10.1145/3132724 >. | DOI 10.1145/3132724
[22] Searle, J. (1984): Mysl, mozek a věda, Mladá fronta, 1994.
[23] Schubbach, A. (2019): Judging machines: philosophical aspects of deep learning, Synthese 198, 1807–1827, dostupné z: < https://doi-org.ezproxy.muni.cz/10.1007/s11229-019-02167-z >.
[24] Silver, D. – Huang, A. – Maddison, C. et al. (2016): Mastering the game of Go with deep neural networks and tree search, Nature 529, 484–489, dostupné z: < https:doi-org.ezproxy.muni.cz/10.1038/nature16961 >.
[25] Slonim, N. – Bilu, Y. – Alzate, C. et al. (2021): An autonomous debating system, Nature 591, 379–384, dostupné z: < https:doi-org.ezproxy.muni.cz/10.1038/s41586-021-03215-w >.
[26] Tan C. et al. (2018): A Survey on Deep Transfer Learning, in Kůrková V. et al. (eds.) Artificial Neural Networks and Machine Learning – ICANN 2018. Lecture Notes in Computer Science 11141, 270–279, dostupné z: < https://doi-org.ezproxy.muni.cz/10.1007/978-3-030-01424-7_27 >.
[27] Toledo, A. et al. (2019): Automatic Argument Quality Assessment – New Datasets and Methods, Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP), D19-1564, 5625–5635, dostupné z: < https://doi-org.ezproxy.muni.cz/10.18653/v1/D19-1564 >.
[28] Turing, A. (1950): Počítacie stroje a inteligencia, Medzinárodná účastnícka spoločnosť Bradlo, 1992.
[29] Turing, A. (1950): Computing Machinery and Intelligence. Mind 59(236), 433–460, dostupné z: < https://doi-org.ezproxy.muni.cz/10.1093/mind/LIX.236.433 >.
[30] Vinyals, O. et al. (2016): Matching Networks for One Shot Learning, Advances in Neural Information Processing Systems 29, 1–12, dostupné z: < https://arxiv.org/abs/1606.04080 >.
[31] Yamins, D. – DiCarlo, J. (2016): Using goal-driven deep learning models to understand sensory cortex, Nat Neurosci 19, 356–365, dostupné z: < https://doi-org.ezproxy.muni.cz/10.1038/nn.4244 > | DOI 10.1038/nn.4244
[32] Zhou, Z-H. (2019): Abductive learning: towards bridging machine learning and logical reasoning, Sci. China Inf. Sci. 62(76101), dostupné z: < https://doi-org.ezproxy.muni.cz/10.1007/s11432-018-9801-4 >.