Разновидности прозрачности: изучение свободы действий в системах ИИ
Журнал «KANT: Social Sciences & Humanities Series» №1(9) 2022 [стр. 35-51]
DOI: 10.24923/2305-8757.2022-9.4
Ключевые слова: прозрачность; этика ИИ; агентство; философия техники; философия разума; философия ИИ.
* Эта работа финансируется за счет национальных фондов через FCT – Фонд науки и техники в рамках проекта UIDB/00183/2020. Роберт В. Клоуз: Эта работа финансируется национальными фондами через FCT – Фонд Науки и Техники, научный стимулирующий грант FCT DL 57/2016/CP1453/CT0021. Пол Смарт: Эта работа поддерживается Исследовательским советом по инженерным и физическим наукам Великобритании (EPSRC) в рамках Национального центра передового опыта PETRAS в области кибербезопасности систем IoT под номером гранта EP/S035362/1.
Системы искусственного интеллекта играют все более важную роль в формировании и регулировании жизни миллионов людей по всему миру. Призывы к большей прозрачности таких систем были широко распространены. Однако существует значительная неясность в отношении того, что на самом деле означает "прозрачность" и, следовательно, что может повлечь за собой повышение прозрачности. В то время как, согласно некоторым дискуссиям, прозрачность требует видеть сквозь предмет или устройство, широко распространенные призывы к прозрачности подразумевают заглянуть внутрь различных аспектов систем ИИ. Эти два понятия явно противоречат друг другу и присутствуют в двух оживленных, но в значительной степени несвязанных дебатах. В этой статье мы стремимся дополнительно проанализировать, что влекут за собой эти призывы к прозрачности, и при этом прояснить виды прозрачности, которые мы должны ожидать от систем ИИ. Мы делаем это, предлагая таксономию, которая классифицирует различные понятия прозрачности. После тщательного изучения различных разновидностей прозрачности мы показываем, как эта таксономия может помочь нам ориентироваться в различных областях взаимодействия человека и технологии, и более полезно обсуждать взаимосвязь между технологической прозрачностью и человеческой деятельностью. В заключение мы утверждаем, что все эти различные понятия прозрачности следует учитывать при разработке более этически адекватных систем ИИ.
Системы искусственного интеллекта играют все более важную роль в формировании и регулировании жизни миллионов людей по всему миру. Призывы к большей прозрачности таких систем были широко распространены. Однако существует значительная неясность в отношении того, что на самом деле означает "прозрачность" и, следовательно, что может повлечь за собой повышение прозрачности. В то время как, согласно некоторым дискуссиям, прозрачность требует видеть сквозь предмет или устройство, широко распространенные призывы к прозрачности подразумевают заглянуть внутрь различных аспектов систем ИИ. Эти два понятия явно противоречат друг другу и присутствуют в двух оживленных, но в значительной степени несвязанных дебатах. В этой статье мы стремимся дополнительно проанализировать, что влекут за собой эти призывы к прозрачности, и при этом прояснить виды прозрачности, которые мы должны ожидать от систем ИИ. Мы делаем это, предлагая таксономию, которая классифицирует различные понятия прозрачности. После тщательного изучения различных разновидностей прозрачности мы показываем, как эта таксономия может помочь нам ориентироваться в различных областях взаимодействия человека и технологии, и более полезно обсуждать взаимосвязь между технологической прозрачностью и человеческой деятельностью. В заключение мы утверждаем, что все эти различные понятия прозрачности следует учитывать при разработке более этически адекватных систем ИИ.
Литература:
1. AI HLEG (High-Level Expert Group on Artificial Intelligence) (2019) Ethics Guidelines for Trustworthy AI. European Commission, Brussels, Belgium. https://digital-strategy.ec.europa.eu/en/library/ethics-guidelines-trustworthy-ai
2. Andrada G (2020) Transparency and the phenomenology of extended cognition. LÍMITE Interdiscipl J Philos Psychol 15(20):1–17
3. Andrada G (2021) Mind the notebook. Synthese 198:4689–4708
4. Bostrom N (2014) Superintelligence: paths, dangers, strategies. Oxford University Press, Oxford
5. Bratman ME (2000) Reflection, planning, and temporally extended agency. Philos Rev 109(1):35–61
6. Bucher T (2012) Want to be on the top? Algorithmic power and the threat of invisibility on Facebook. New Media Soc 14(7):1164–1180
7. Carter JA (2020) Intellectual autonomy, epistemic dependence and cognitive enhancement. Synthese 197(7):2937–2961
Article Google Scholar
8. Clark A (2008) Supersizing the mind: embodiment, action, and cognitive extension. Oxford University Press, New York
9. Clark A, Chalmers D (1998) The extended mind. Analysis 58(1):7–19
10. Clowes RW (2015) Thinking in the cloud: the cognitive incorporation of cloud-based technology. Philos Technol 28(2):261–296. https://doi.org/10.1007/s13347-014-0153-z
11. Clowes RW (2019a) Immaterial engagement: Human agency and the cognitive ecology of the Internet. Phenomenol Cogn Sci 18(1):259–279
12. Clowes RW (2019b) Screen reading and the creation of new cognitive ecologies. AI Soc 34:705–720
13. Clowes RW (2020) The internet extended person: exoself or doppelganger? LÍMITE Interdiscipl J Philos Psychol 15(22):1–23
14. Coeckelbergh M (2020) AI ethics. MIT Press, Cambridge
15. Cristianini N, Scantamburlo T (2020) On social machines for algorithmic regulation. AI Soc 35:645–662
16. de Fine Licht K, de Fine Licht J (2020) Artificial intelligence, transparency, and public decision-making. AI Soc 35(4):917–926
17. Diakopoulos N (2020) Transparency. In: Dubber MD, Pasquale F, Das S (eds) The oxford handbook of ethics of AI. Oxford University Press, New York, pp 197–213
18. Dreyfus SE, Dreyfus HL (1980) A five-stage model of the mental activities involved in directed skill acquisition. In: Operations Research Center, University of California, Berkeley, California
19. Ferreira FGDC, Gandomi AH, Cardoso RTN (2021) Artificial intelligence applied to stock market trading: a review. IEEE Access 9:30898–30917
20. Floridi L, Cowls J, Beltrametti M, Chatila R, Chazerand P, Dignum V, Luetge C, Madelin R, Pagallo U, Rossi F (2018) AI4People—an ethical framework for a good AI society: opportunities, risks, principles, and recommendations. Mind Mach 28(4):689–707
21. Gallagher S (2005) How the body shapes the mind. Oxford University Press, Oxford
22. Gillett AJ, Heersmink R (2019) How navigation systems transform epistemic virtues: knowledge, issues and solutions. Cogn Syst Res 56:36–49
23. Heersmink R (2013) A taxonomy of cognitive artifacts: function, information, and categories. Rev Philos Psychol 4(3):465–481
24. Heersmink R (2015) Dimensions of integration in embedded and extended cognitive systems. Phenomenol Cogn Sci 14(3):577–598
25. Heersmink R, Sutton J (2020) Cognition and the web: extended, transactive, or scaffolded? Erkenntnis 85:139–164
26. Heidegger M (1927) Being and time. Basil Blackwell, Oxford
27. Lupton, D. (2016) Digital health technologies and digital data: new ways of monitoring, measuring and commodifying human bodies. In: Olleros FX, Zhegu M (eds) Research handbook on digital transformations. Edward Elgar Publishing Ltd., Cheltenham
28. Maravita A, Iriki A (2004) Tools for the body (schema). Trends Cogn Sci 8(2):79–86
29. Merleau-Ponty M (1945) Phenomenology of Perception. Routledge Press, London
30. Müller VC (2020) Ethics of artificial intelligence and robotics. In: Zalta EN (ed) The stanford encyclopedia of philosophy (Fall 2020 ed.). Stanford University, Stanford, California, USA. https://plato.stanford.edu/archives/fall2020/entries/ethics-ai/
31. Nguyen CT (2021) Transparency is surveillance. Philos Phenomenol Res. https://doi.org/10.1111/phpr.12823
32. O’Neill O (2020) Questioning Trust. In: Simon J (ed) The routledge handbook of trust and philosophy. Routledge, New York, pp 17–27
33. Russell SJ (2019) Human compatible: AI and the problem of control. Viking Press, New York
34. Smart PR, Heersmink R, Clowes RW (2017) The cognitive ecology of the internet. In: Cowley SJ, Vallée-Tourangeau F (eds) Cognition beyond the brain: computation, interactivity and human artifice (2nd ed, pp 251–282). Springer International Publishing, Cham, Switzerland
35. Turilli M, Floridi L (2009) The ethics of information transparency. Ethics Inf Technol 11(2):105–112
36. Walmsley J (2020) Artificial intelligence and the value of transparency. AI Soc 36(2):585–595
37. Wang F-Y (2008) Toward a revolution in transportation operations: AI for complex systems. IEEE Intell Syst 23(6):8–13
38. Weller A (2019) Transparency: motivations and challenges. In: Samek W, Montavon G, Vedaldi A, Hansen LK, Müller K-R (eds) Explainable AI: interpreting, explaining and visualizing deep learning (Vol 11700, pp 23–40). Springer, Cham, Switzerland
39. Wheeler M (2019) The reappearing tool: transparency, smart technology, and the extended mind. AI Soc 34(4):857–866
40. Zednik C (2021) Solving the black box problem: a normative framework for explainable artificial intelligence. Philos Technol 34:265–288.
1. AI HLEG (High-Level Expert Group on Artificial Intelligence) (2019) Ethics Guidelines for Trustworthy AI. European Commission, Brussels, Belgium. https://digital-strategy.ec.europa.eu/en/library/ethics-guidelines-trustworthy-ai
2. Andrada G (2020) Transparency and the phenomenology of extended cognition. LÍMITE Interdiscipl J Philos Psychol 15(20):1–17
3. Andrada G (2021) Mind the notebook. Synthese 198:4689–4708
4. Bostrom N (2014) Superintelligence: paths, dangers, strategies. Oxford University Press, Oxford
5. Bratman ME (2000) Reflection, planning, and temporally extended agency. Philos Rev 109(1):35–61
6. Bucher T (2012) Want to be on the top? Algorithmic power and the threat of invisibility on Facebook. New Media Soc 14(7):1164–1180
7. Carter JA (2020) Intellectual autonomy, epistemic dependence and cognitive enhancement. Synthese 197(7):2937–2961
Article Google Scholar
8. Clark A (2008) Supersizing the mind: embodiment, action, and cognitive extension. Oxford University Press, New York
9. Clark A, Chalmers D (1998) The extended mind. Analysis 58(1):7–19
10. Clowes RW (2015) Thinking in the cloud: the cognitive incorporation of cloud-based technology. Philos Technol 28(2):261–296. https://doi.org/10.1007/s13347-014-0153-z
11. Clowes RW (2019a) Immaterial engagement: Human agency and the cognitive ecology of the Internet. Phenomenol Cogn Sci 18(1):259–279
12. Clowes RW (2019b) Screen reading and the creation of new cognitive ecologies. AI Soc 34:705–720
13. Clowes RW (2020) The internet extended person: exoself or doppelganger? LÍMITE Interdiscipl J Philos Psychol 15(22):1–23
14. Coeckelbergh M (2020) AI ethics. MIT Press, Cambridge
15. Cristianini N, Scantamburlo T (2020) On social machines for algorithmic regulation. AI Soc 35:645–662
16. de Fine Licht K, de Fine Licht J (2020) Artificial intelligence, transparency, and public decision-making. AI Soc 35(4):917–926
17. Diakopoulos N (2020) Transparency. In: Dubber MD, Pasquale F, Das S (eds) The oxford handbook of ethics of AI. Oxford University Press, New York, pp 197–213
18. Dreyfus SE, Dreyfus HL (1980) A five-stage model of the mental activities involved in directed skill acquisition. In: Operations Research Center, University of California, Berkeley, California
19. Ferreira FGDC, Gandomi AH, Cardoso RTN (2021) Artificial intelligence applied to stock market trading: a review. IEEE Access 9:30898–30917
20. Floridi L, Cowls J, Beltrametti M, Chatila R, Chazerand P, Dignum V, Luetge C, Madelin R, Pagallo U, Rossi F (2018) AI4People—an ethical framework for a good AI society: opportunities, risks, principles, and recommendations. Mind Mach 28(4):689–707
21. Gallagher S (2005) How the body shapes the mind. Oxford University Press, Oxford
22. Gillett AJ, Heersmink R (2019) How navigation systems transform epistemic virtues: knowledge, issues and solutions. Cogn Syst Res 56:36–49
23. Heersmink R (2013) A taxonomy of cognitive artifacts: function, information, and categories. Rev Philos Psychol 4(3):465–481
24. Heersmink R (2015) Dimensions of integration in embedded and extended cognitive systems. Phenomenol Cogn Sci 14(3):577–598
25. Heersmink R, Sutton J (2020) Cognition and the web: extended, transactive, or scaffolded? Erkenntnis 85:139–164
26. Heidegger M (1927) Being and time. Basil Blackwell, Oxford
27. Lupton, D. (2016) Digital health technologies and digital data: new ways of monitoring, measuring and commodifying human bodies. In: Olleros FX, Zhegu M (eds) Research handbook on digital transformations. Edward Elgar Publishing Ltd., Cheltenham
28. Maravita A, Iriki A (2004) Tools for the body (schema). Trends Cogn Sci 8(2):79–86
29. Merleau-Ponty M (1945) Phenomenology of Perception. Routledge Press, London
30. Müller VC (2020) Ethics of artificial intelligence and robotics. In: Zalta EN (ed) The stanford encyclopedia of philosophy (Fall 2020 ed.). Stanford University, Stanford, California, USA. https://plato.stanford.edu/archives/fall2020/entries/ethics-ai/
31. Nguyen CT (2021) Transparency is surveillance. Philos Phenomenol Res. https://doi.org/10.1111/phpr.12823
32. O’Neill O (2020) Questioning Trust. In: Simon J (ed) The routledge handbook of trust and philosophy. Routledge, New York, pp 17–27
33. Russell SJ (2019) Human compatible: AI and the problem of control. Viking Press, New York
34. Smart PR, Heersmink R, Clowes RW (2017) The cognitive ecology of the internet. In: Cowley SJ, Vallée-Tourangeau F (eds) Cognition beyond the brain: computation, interactivity and human artifice (2nd ed, pp 251–282). Springer International Publishing, Cham, Switzerland
35. Turilli M, Floridi L (2009) The ethics of information transparency. Ethics Inf Technol 11(2):105–112
36. Walmsley J (2020) Artificial intelligence and the value of transparency. AI Soc 36(2):585–595
37. Wang F-Y (2008) Toward a revolution in transportation operations: AI for complex systems. IEEE Intell Syst 23(6):8–13
38. Weller A (2019) Transparency: motivations and challenges. In: Samek W, Montavon G, Vedaldi A, Hansen LK, Müller K-R (eds) Explainable AI: interpreting, explaining and visualizing deep learning (Vol 11700, pp 23–40). Springer, Cham, Switzerland
39. Wheeler M (2019) The reappearing tool: transparency, smart technology, and the extended mind. AI Soc 34(4):857–866
40. Zednik C (2021) Solving the black box problem: a normative framework for explainable artificial intelligence. Philos Technol 34:265–288.
Varieties of transparency: exploring agency within AI systems
Keywords: transparency; AI ethics; agency; philosophy of technology; philosophy of mind; philosophy of AI.
* This work is funded by national funds through the FCT Foundation for Science and Technology within the framework of the UIDB/00183/2020 project. Robert W. Close: This work is funded by national foundations through the FCT Foundation for Science and Technology, a scientific incentive grant FCT DL 57/2016/CP1453/CT0021. Paul Smart: This work is supported by the UK Engineering and Physical Sciences Research Council (EPSRC) as part of the PETRAS National Center of Excellence for Cybersecurity of IoT Systems under grant number EP/S035362/1.
AI systems play an increasingly important role in shaping and regulating the lives of millions of human beings across the world. Calls for greater transparency from such systems have been widespread. However, there is considerable ambiguity concerning what "transparency" actually means, and therefore, what greater transparency might entail. While, according to some debates, transparency requires seeing through the artefact or device, widespread calls for transparency imply seeing into different aspects of AI systems. These two notions are in apparent tension with each other, and they are present in two lively but largely disconnected debates. In this paper, we aim to further analyse what these calls for transparency entail, and in so doing, clarify the sorts of transparency that we should want from AI systems. We do so by offering a taxonomy that classifies different notions of transparency. After a careful exploration of the different varieties of transparency, we show how this taxonomy can help us to navigate various domains of human-technology interactions, and more usefully discuss the relationship between technological transparency and human agency. We conclude by arguing that all of these different notions of transparency should be taken into account when designing more ethically adequate AI systems.
Andrada, G., Clowes, R.W. & Smart, P.R. Varieties of transparency: exploring agency within AI systems. AI & Society (2022).
AI systems play an increasingly important role in shaping and regulating the lives of millions of human beings across the world. Calls for greater transparency from such systems have been widespread. However, there is considerable ambiguity concerning what "transparency" actually means, and therefore, what greater transparency might entail. While, according to some debates, transparency requires seeing through the artefact or device, widespread calls for transparency imply seeing into different aspects of AI systems. These two notions are in apparent tension with each other, and they are present in two lively but largely disconnected debates. In this paper, we aim to further analyse what these calls for transparency entail, and in so doing, clarify the sorts of transparency that we should want from AI systems. We do so by offering a taxonomy that classifies different notions of transparency. After a careful exploration of the different varieties of transparency, we show how this taxonomy can help us to navigate various domains of human-technology interactions, and more usefully discuss the relationship between technological transparency and human agency. We conclude by arguing that all of these different notions of transparency should be taken into account when designing more ethically adequate AI systems.
Andrada, G., Clowes, R.W. & Smart, P.R. Varieties of transparency: exploring agency within AI systems. AI & Society (2022).