Papers
Topics
Authors
Recent
Search
2000 character limit reached

Inductive Models for Artificial Intelligence Systems are Insufficient without Good Explanations

Published 17 Jan 2024 in cs.LG and cs.AI | (2401.09011v1)

Abstract: This paper discusses the limitations of ML, particularly deep artificial neural networks (ANNs), which are effective at approximating complex functions but often lack transparency and explanatory power. It highlights the `problem of induction' : the philosophical issue that past observations may not necessarily predict future events, a challenge that ML models face when encountering new, unseen data. The paper argues for the importance of not just making predictions but also providing good explanations, a feature that current models often fail to deliver. It suggests that for AI to progress, we must seek models that offer insights and explanations, not just predictions.

Authors (1)
Definition Search Book Streamline Icon: https://streamlinehq.com
References (62)
  1. First m87 event horizon telescope results. iv. imaging the central supermassive black hole. The Astrophysical Journal Letters, 875(1):L4, 2019.
  2. Afra Alishahi. Computational modeling of human language acquisition. Morgan & Claypool Publishers, 2010.
  3. Francis Bacon. Novum organum. Clarendon press, 1878.
  4. Neural machine translation by jointly learning to align and translate. arXiv preprint arXiv:1409.0473, 2014.
  5. A human-centered evaluation of a deep learning system deployed in clinics for the detection of diabetic retinopathy. In Proceedings of the 2020 CHI conference on human factors in computing systems, pages 1–12, 2020.
  6. Sparks of artificial general intelligence: Early experiments with gpt-4. arXiv preprint arXiv:2303.12712, 2023.
  7. Walter Kaufmann Bühler. Gauss: a biographical study. Springer Science & Business Media, 2012.
  8. Support-vector networks. Machine learning, 20:273–297, 1995.
  9. David R Cox. The regression analysis of binary sequences. Journal of the Royal Statistical Society Series B: Statistical Methodology, 20(2):215–232, 1958.
  10. George Cybenko. Approximation by superpositions of a sigmoidal function. Mathematics of control, signals and systems, 2(4):303–314, 1989.
  11. David Deutsch. The beginning of infinity: Explanations that transform the world. penguin uK, 2011.
  12. Towards a rigorous science of interpretable machine learning. arXiv preprint arXiv:1702.08608, 2017.
  13. Ix. a determination of the deflection of light by the sun’s gravitational field, from observations made at the total eclipse of may 29, 1919. Philosophical Transactions of the Royal Society of London. Series A, Containing Papers of a Mathematical or Physical Character, 220(571-581):291–333, 1920.
  14. Albert Einstein. Die grundlage der allgemeinen relativitätstheorie, volume 49. JA Barth, 1922.
  15. Daniel C Elton. Self-explaining AI as an alternative to interpretable AI. In Artificial General Intelligence: 13th International Conference, AGI 2020, St. Petersburg, Russia, September 16–19, 2020, Proceedings 13, pages 95–106. Springer, 2020.
  16. Daniel C Elton. Applying deutsch’s concept of good explanations to artificial intelligence and neuroscience–an initial exploration. Cognitive Systems Research, 67:9–17, 2021.
  17. Francis Galton. Regression towards mediocrity in hereditary stature. The Journal of the Anthropological Institute of Great Britain and Ireland, 15:246–263, 1886.
  18. Carl Friedrich Gauss. Theoria motus corporum coelestium in sectionibus conicis solem ambientium, volume 7. FA Perthes, 1877.
  19. Generative adversarial nets. Advances in neural information processing systems, 27, 2014.
  20. A survey of methods for explaining black box models. ACM computing surveys (CSUR), 51(5):1–42, 2018.
  21. Stellar interiors: physical principles, structure, and evolution. Springer Science & Business Media, 2012.
  22. Direct fit to nature: an evolutionary perspective on biological and artificial neural networks. Neuron, 105(3):416–434, 2020.
  23. John Haugeland. Artificial intelligence: The very idea. MIT press, 1989.
  24. Denoising diffusion probabilistic models. Advances in neural information processing systems, 33:6840–6851, 2020.
  25. Long short-term memory. Neural computation, 9(8):1735–1780, 1997.
  26. Kurt Hornik. Approximation capabilities of multilayer feedforward networks. Neural networks, 4(2):251–257, 1991.
  27. David Hume. An enquiry concerning human understanding. In Seven masterpieces of philosophy, pages 183–276. Routledge, 2016.
  28. Adversarial examples are not bugs, they are features. Advances in neural information processing systems, 32, 2019.
  29. Attention is not explanation. arXiv preprint arXiv:1902.10186, 2019.
  30. Qualia. Current Biology, 22(10):R392–R396, 2012.
  31. Schema networks: Zero-shot transfer with a generative causal model of intuitive physics. In International conference on machine learning, pages 1809–1818. PMLR, 2017.
  32. Backpropagation applied to handwritten zip code recognition. Neural computation, 1(4):541–551, 1989.
  33. Adrien Marie Legendre. Nouvelles méthodes pour la détermination des orbites des comètes: avec un supplément contenant divers perfectionnemens de ces méthodes et leur application aux deux comètes de 1805. Courcier, 1806.
  34. Cyc: toward programs with common sense. Communications of the ACM, 33(8):30–49, 1990.
  35. Autonomous driving: technical, legal and social aspects. Springer Nature, 2016.
  36. A proposal for the dartmouth summer research project on artificial intelligence, august 31, 1955. AI magazine, 27(4):12–12, 2006.
  37. A logical calculus of the ideas immanent in nervous activity. The bulletin of mathematical biophysics, 5:115–133, 1943.
  38. Hans P Moravec. The stanford cart and the cmu rover. Proceedings of the IEEE, 71(7):872–884, 1983.
  39. Computer science as empirical inquiry: Symbols and search. In ACM Turing award lectures, page 1975. 2007.
  40. Induction, popper, and machine learning. arXiv preprint arXiv:2110.00840, 2021.
  41. Cathy O’neil. Weapons of math destruction: How big data increases inequality and threatens democracy. Crown, 2017.
  42. Karl Pearson. Contributions to the mathematical theory of evolution. Philosophical Transactions of the Royal Society of London. A, 185:71–110, 1894.
  43. Karl Pearson. Liii. on lines and planes of closest fit to systems of points in space. The London, Edinburgh, and Dublin philosophical magazine and journal of science, 2(11):559–572, 1901.
  44. Steven Piantadosi. Modern language models refute chomsky’s approach to language. Lingbuzz Preprint, lingbuzz, 7180, 2023.
  45. Karl Raimund Popper et al. Objective knowledge: An evolutionary approach, volume 49. Clarendon press Oxford, 1979.
  46. Karl Popper. The logic of scientific discovery. Routledge, 2005.
  47. Karl Popper. Conjectures and refutations: The growth of scientific knowledge. routledge, 2014.
  48. J. Ross Quinlan. Induction of decision trees. Machine learning, 1:81–106, 1986.
  49. Frank Rosenblatt. The perceptron: a probabilistic model for information storage and organization in the brain. Psychological review, 65(6):386, 1958.
  50. Cynthia Rudin. Stop explaining black box machine learning models for high stakes decisions and use interpretable models instead. Nature machine intelligence, 1(5):206–215, 2019.
  51. Learning representations by back-propagating errors. nature, 323(6088):533–536, 1986.
  52. S Russell and P Norvig. Artificial Intelligence: A Modern Approach. Pearson, 2016.
  53. Our sun. iii. present and future. Astrophysical Journal v. 418, p. 457, 418:457, 1993.
  54. Fairness and abstraction in sociotechnical systems. In Proceedings of the conference on fairness, accountability, and transparency, pages 59–68, 2019.
  55. Intriguing properties of neural networks. arXiv preprint arXiv:1312.6199, 2013.
  56. Judith Jarvis Thomson. The trolley problem. Yale LJ, 94:1395, 1984.
  57. Attention is all you need. Advances in neural information processing systems, 30, 2017.
  58. Attention is not not explanation. arXiv preprint arXiv:1908.04626, 2019.
  59. Terry Winograd. Understanding natural language. Cognitive psychology, 3(1):1–191, 1972.
  60. Stephen Wolfram. Wolfram alpha documentation timeline. https://www.wolframalpha.com/docs/timeline, 2023. Accessed: 2023-11-07.
  61. Understanding deep learning (still) requires rethinking generalization. Communications of the ACM, 64(3):107–115, 2021.
  62. How language model hallucinations can snowball. arXiv preprint arXiv:2305.13534, 2023.

Summary

No one has generated a summary of this paper yet.

Sign Up to Summarize

Paper to Video (Beta)

No one has generated a video about this paper yet.

Sign Up to Generate All Videos Subscribe on YouTube

Whiteboard

No one has generated a whiteboard explanation for this paper yet.

Sign Up to Generate

Open Problems

We haven't generated a list of open problems mentioned in this paper yet.

Generate Now

Continue Learning

We haven't generated follow-up questions for this paper yet.

Generate Now

Collections

Sign up for free to add this paper to one or more collections.

Sign Up

Tweets

Sign up for free to view the 1 tweet with 1 like about this paper.

Sign Up for Free