Abstract
In recent years, the implementation of meaningfully controllable advanced intelligent systems whose goals are aligned with ethical values as specified by human entities emerged as key subject of investigation of international relevance across diverse AI-related research areas. In this paper, we present a novel transdisciplinary and Systems Engineering oriented approach denoted “orthogonality-based disentanglement” which jointly tackles both the thereby underlying control problem and value alignment problem while unraveling the corresponding responsibilities of different stakeholders based on the distinction of two orthogonal axes assigned to the problem-solving ability of these intelligent systems on the one hand and to the ethical abilities they exhibit based on quantitatively encoded human values on the other hand. Moreover, we introduce the notion of explicitly formulated ethical goal functions ideally encoding what humans should want and exemplify a possible class of “self-aware” intelligent systems with the capability to reliably adhere to these human-defined goal functions. Beyond that, we discuss an attainable transformative socio-technological feedback-loop that could result out of the introduced orthogonality-based disentanglement approach and briefly elaborate on how the framework additionally provides valuable hints with regard to the coordination subtask in AI Safety. Finally, we point out remaining crucial challenges as incentive for future work.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Similar content being viewed by others
References
Aliman, N.-M., Kester, L.: Hybrid strategies towards safe self-aware superintelligent systems. In: Iklé, M., Franz, A., Rzepka, R., Goertzel, B. (eds.) AGI 2018. LNCS (LNAI), vol. 10999, pp. 1–11. Springer, Cham (2018). https://doi.org/10.1007/978-3-319-97676-1_1
Aliman, N.-M., Kester, L.: Augmented Utilitarianism for AGI Safety. In: Hammer, P. Agrawal, P., Goertzel, B., Iklé, M. (eds.) AGI 2019. LNAI, vol. 11654, pp. 11–21. Springer, Cham (2019)
Aliman, N.M., Kester, L.: Transformative AI governance and AI-Empowered ethical enhancement through preemptive simulations. Delphi - Interdisc. Rev. Emerg. Technol. 2(1), 23–29 (2019)
Armstrong, S.: General purpose intelligence: arguing the orthogonality thesis. Anal. Metaphys. 12, 68–84 (2013)
Bostrom, N.: The superintelligent will: motivation and instrumental rationality in advanced artificial agents. Mind. Mach. 22(2), 71–85 (2012)
Brundage, M., et al.: The Malicious Use of Artificial Intelligence: Forecasting, Prevention, and Mitigation. arXiv preprint arXiv:1802.07228 (2018)
Eckersley, P.: Impossibility and Uncertainty Theorems in AI Value Alignment (or why your AGI should not have a utility function). CoRR abs/1901.00064 (2018)
Elands, P., Huizing, A., Kester, L., Oggero, S., Peeters, M.: Governing Ethical and Effective Behaviour of Intelligent Systems. Military spectator (2019, to appear)
Everitt, T., Lea, G., Hutter, M.: AGI Safety Literature Review. arXiv preprint arXiv:1805.01109 (2018)
Goertzel, B.: Infusing advanced AGIs with human-like value systems: two theses. J. Evol. Technol. 26(1), 50–72 (2016)
Harris, S.: The Moral Landscape: How Science can Determine Human Values. Simon and Schuster, New York (2011)
Hoekstra, R., Breuker, J., Di Bello, M., Boer, A., et al.: The LKIF core ontology of basic legal concepts. LOAIT 321, 43–63 (2007)
Kester, L., Ditzel, M.: Maximising effectiveness of distributed mobile observation systems in dynamic situations. In: 2014 17th International Conference on Information Fusion (FUSION), pp. 1–8. IEEE (2014)
Kester, L.J.H.M., van Willigen, W.H., Jongh, J.D.: Critical headway estimation under uncertainty and non-ideal communication conditions. In: 17th International IEEE Conference on Intelligent Transportation Systems (ITSC) pp. 320–327 (2014)
Korteling, J.E., Brouwer, A.M., Toet, A.: A neural network framework for cognitive bias. Frontiers in Psychol. 9, 1561 (2018)
Leike, J., Krueger, D., Everitt, T., Martic, M., Maini, V., Legg, S.: Scalable agent alignment via reward modeling: a research direction. arXiv preprint arXiv:1811.07871 (2018)
Pistono, F., Yampolskiy, R.V.: Unethical research: how to create a malevolent artificial intelligence. In: 25th International Joint Conference on Artificial Intelligence (IJCAI-2016). Ethics for Artificial Intelligence Workshop (AI-Ethics-2016) (2016)
Poel, I.: Translating values into design requirements. In: Michelfelder, D.P., McCarthy, N., Goldberg, D.E. (eds.) Philosophy and Engineering: Reflections on Practice, Principles and Process. PET, vol. 15, pp. 253–266. Springer, Dordrecht (2013). https://doi.org/10.1007/978-94-007-7762-0_20
Russell, S., Dewey, D., Tegmark, M.: Research priorities for robust and beneficial artificial intelligence. AI Mag. 36(4), 105–114 (2015)
Sezer, O., Gino, F., Bazerman, M.H.: Ethical blind spots: explaining unintentional unethical behavior. Current Opin. Psychol. 6, 77–81 (2015)
Thorisson, K.R.: A new constructivist AI: from manual methods to self-constructive systems. Theoretical Foundations of Artificial General Intelligence. Atlantis Thinking Machines, vol. 4, pp. 145–171. Springer, Atlantis Press, Paris (2012). https://doi.org/10.2991/978-94-91216-62-6_9
Tomsett, R., et al.: Why the failure? how adversarial examples can provide insights for interpretable machine learning. In: 2018 21st International Conference on Information Fusion (FUSION), pp. 838–845. IEEE (2018)
Werkhoven, P., Kester, L., Neerincx, M.: Telling autonomous systems what to do. In: Proceedings of the 36th European Conference on Cognitive Ergonomics, p. 2. ACM (2018)
Yudkowsky, E.: The AI alignment problem: why it is hard, and where to start. Symbolic Systems Distinguished Speaker (2016)
Yudkowsky, E.: Coherent extrapolated volition. In: Singularity Institute for Artificial Intelligence (2004)
Ziesche, S.: Potential synergies between the united nations sustainable development goals and the value loading problem in artificial intelligence. Maldives National J. Res. 6, 47 (2018)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2019 Springer Nature Switzerland AG
About this paper
Cite this paper
Aliman, NM., Kester, L., Werkhoven, P., Yampolskiy, R. (2019). Orthogonality-Based Disentanglement of Responsibilities for Ethical Intelligent Systems. In: Hammer, P., Agrawal, P., Goertzel, B., Iklé, M. (eds) Artificial General Intelligence. AGI 2019. Lecture Notes in Computer Science(), vol 11654. Springer, Cham. https://doi.org/10.1007/978-3-030-27005-6_3
Download citation
DOI: https://doi.org/10.1007/978-3-030-27005-6_3
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-27004-9
Online ISBN: 978-3-030-27005-6
eBook Packages: Computer ScienceComputer Science (R0)