淺論AI風險預測的規範性爭議

A Preliminary Study of Normative Issues of AI Prediction

10.7015/JEAS.202006_50(2).0004

洪子偉(Tzu-Wei Hung)

規範性 ； 歸納法 ； 隱私 ； 人工智慧 ； 預測技術 ； normativity ； induction ； privacy ； artificial intelligence ； predictive technology

歐美研究

50卷2期（2020 / 06 / 01）

207 - 229

繁體中文

AI預測技術深具潛力的其中一個應用，在於分析過去資料以預防極端氣候的災害。但當預測對象從自然環境變成人類本身，爭議隨之產生。本文旨在探討以人類資料作為風險預測之規範性爭議，並主張：（一）AI的不可解釋性，並非因其無法提供機械式步驟，而是人類的認知限制無法對數量龐大的步驟賦予意義。（二）AI的歸納法、黑箱等問題在大腦上也會遇到，兩者的差異是程度上而非種類上的。（三）必然性與事實性條件並無法一致地排除極端案例，卻又不排除既有法律或社會規範。（四）自主性原則之優點在確保權責相符、避免喪失人類能力、降低AI發展的社會阻力。

This paper focuses on normative aspects of AI prediction-that is, technologies used to predict the future through analyses of big data concerning the past. While this technology seems promising in forecasting extreme weather or rehabilitating endangered wildlife, it is controversial when applied to human beings, e.g., an Israeli company is using AI prediction to identify possible terrorists, and China's government to locate potential dissidents. This paper explores some of the normative issues and argues: (1) AI-derived conclusions are inexplicable not because machines fail to provide mechanical steps, but because our limited cognitive power cannot assign meaning to the, probably billions of, steps, and thus we fail to understand the conclusions reached by AI; (2) while AI is considered to have an inductive problem, to be a black box, and to have other epistemological issues, these worries apply to the human brain as well. AI and the human brain are different in degree rather than type; (3) the necessity argument and the reality condition cannot be used to exclude radical cases (e.g., China's social credit system) without excluding existing laws or social norms; and (4) the principle of autonomy has advantages, which include balancing power and responsibility, and reduces public distrust.

1. Brown, H.,Friston, K. J.(2012).Free-energy and illusions: The cornsweet effect.Frontiers in Psychology,3,Article 43.
2. Burgin, M.(2004).Algorithmic complexity of recursive and inductive algorithms.Theoretical Computer Science,317(1/2/3),31-60.
3. Burgin, M.(2005).Super-recursive algorithms.New York:Springer.
4. Church, A. (1935). Abstract no. 204. Bulletin of the American Mathematical Society, 41: 332-333.
5. Church, A. (1936). An unsolvable problem of elementary number theory. American Journal of Mathematics, 58: 345-363.
6. Clark, A.(2013).Whatever next? Predictive brains, situated agents, and the future of cognitive science.Behavioral and Brain Sciences,36(3),181-204.
7. Dennett, D. C.(1989).The intentional stance.Cambridge, MA:MIT Press.
8. Floridi, L. (2016). Should we be afraid of AI. Aeon Essays. Retrieved from https://aeon.co/essays/true-ai-is-both-logically-possible-and-utterly-implausible
9. Fodor, J. A.(2008).LOT 2: The language of thought revisited.Oxford, UK:Oxford University Press.
10. Hung, T.-W.,Yen, C.-P.(2021).On the person-based predictive policing of AI.Ethics and Information Technology,23(3)
11. Ince, D. C.(Ed.)(1992).Mechanical intelligence: Collected works of A. M. Turing.Edinburgh, UK:Edinburgh University Press.
12. Jones, N.(2017).How machine learning could help to improve climate forecasts.Nature News,548(7668),379.
13. Kleene, S. C.(1967).Mathematical logic.New York:Wiley.
14. Lin, Y.-T.,Hung, T.-W.,Huang, T. L.(2021).Engineering equity: How AI can help reduce the harm of implicit bias.Philosophy & Technology,34
15. Liu, Y.,Racah, E.,Correa, J.,Khosrowshahi, A.,Lavers, D.,Kunkel, K.(2016).,未出版
16. Marr, D.(1982).Vision: A computational investigation into the human representation and processing of visual information.New York:Henry Holt.
17. Mozur, P. (2018, March 2). China presses its internet censorship efforts across the globe. New York Times. Retrieved from https://www.nytimes.com/
18. Murphy, H. (2017, October 9). Why Stanford researchers tried to create a “Gaydar” machine. New York Times. Retrieved from https://www.nytimes.com/
19. Nebehay, S. (2018, August 30). U. N. calls on China to free Uighurs from alleged re-education camps. Reuters World News. Retrieved from https://www.reuters.com/
20. Pearl, J.,Mackenzie, D.(2018).The book of why: The new science of cause and effect.New York:Basic Books.
21. Piccinini, G.,Bahar, S.(2013).Neural computation and the computational theory of cognition.Cognitive Science,37(3),453-488.
22. Rivers, M. (2020, January 3). More than 100 Uyghur graveyards demolished by Chinese authorities, satellite images show. CNN. Retrieved from https://edition.cnn.com
23. Schramm, J. D. (2018, February 19). AI “gaydar” could compromise LGBTQ people’s privacy-and safety. Washington Post. Retrieved from https://www.washingtonpost.com/
24. Sprevak, M.(Ed.),Colombo, M.(Ed.)(2018).Routledge handbook to the computational mind.Oxford, UK:Routledge.
25. Swanson, L. R.(2016).The predictive processing paradigm has roots in Kant.Frontiers in Systems Neuroscience,10,Article 79.
26. Thomas, D. (2018, July 17) The cameras that know if you’re happy – or a threat. BBC News. Retrieved from https://www.bbc.co.uk
27. Tomlinson, S. (2016, May 24). Can you spot a terrorist just by lookingat their face? Daily Mail. Retrieved from https://www.dailymail.co.uk
28. Turing, A. M. (1937b). On computable numbers, with an application to the Entscheidungs problem. Proceedings of the London Mathematical Society, S2, 42: 230-265. https://doi.org/10.1112/plms/s2-42.1.230
29. Turing, A. M. (1937a). Computability and λ-definability. Journal of Symbolic Logic, 2, 4: 153-163. https://doi.org/10.2307/2268280
30. Vincent, J., & Brandom, R. (2018, April 26). Axon launches AI ethics board to study the dangers of facial recognition. The Verge. Retrieved from https://www.theverge.com/
31. Wachter, S.,Mittelstadt, B.,Floridi, L.(2017).Why a right to explanation of automated decision-making does not exist in the general data protection regulation.International Data Privacy Law,7(2),76-99.
32. Yang, G. Z.,Bellingham, J.,Dupont, P. E.,Fischer, P.,Floridi, L.,Full, R.(2018).The grand challenges of science robotics.Science Robotics,3(14),eaar7650.

1. (2023)。智慧醫療器材與醫師之注意義務初探（下）。成大法學，46，185-236。
2. (2024)。新聞媒體對於生成式人工智慧的感知風險、規範與實踐。新聞學研究，160，1-66。