Artificial Intelligence Align with Human Values
Keywords:
Artificial Intelligence, Phenomenological Theories, Human Values, Sustainable Software Development, Machine Learning, AI AutomationAbstract
Artificial Intelligence (AI) should be directed at humane ends. The event of AI has created nice uncertainties of guaranteeing AI alignment with humans’ worth (AI value alignment) through AI operations from style to use. For the needs of addressing this problem, we tend to adopt the phenomenological theories of fabric values and technological mediation to be that starting step for the upcoming future. During this paper, initially, we will discuss the AI value alignment from the relevant AI studies. Second, we concisely gift what are material values and technological mediation and replicate on the AI value alignment through the lenses of those theories. We tend to conclude that a set of finite human worth’s are often outlined and tailored to the stable life tasks that AI systems will be known as upon to accomplish. The AI value alignment can even be fostered between designers and users through technological mediation. Upon that foundation, we tend to propose a collection of common principles to grasp the AI value alignment through phenomenological theories. This paper contributes the distinctive information of phenomenological theories to the discourse on AI alignment with human values.
References
1. Stuart Russell. Aligning AI with human values. Available at: https://arxiv.org/abs/1606.06565.
2. Francesca Rossi, Nicholas Mattei. Towards aligning human and machine values in AI. Available at: https://arxiv.org/abs/1805.02237.
3. Dylan Hadfield -Menell et al. Value alignment in cooperative inverse reinforcement learning . Available at: https://arxiv.org/abs/1606.03137.
4. Andrew Critch, Stuart Armstrong. Value alignment problems in AI ethics. Available at: https://arxiv.org/abs/1706.06311.
5. Nick Bostrom . Aligning superintelligence with human values . Available at: https://nickbostrom.com/superintelligence.pdf.
6. Dylan Hadfield -Menell et al . Foundations for aligning AI with human values . Available at: https://arxiv.org/abs/1711.05042.
7. Paula Boddington . Towards a code of ethics for artificial intelligence . Available at: https://doi.org/10.1007/978-3-319-70284-1_1.
8. Daniel Weinstock . The alignment problem in ethics and political philosophy . Available at: https://link.springer.com/chapter/10.1007/978-3-030-01446-3_3.
9. Nick Bostrom, Eliezer Yudkowsky. Ethical issues in advanced artificial intelligence. Available at: https://www.nickbostrom.com/ethics/ai.pdf.
10. Victoria Krakovna . AI alignment: Why it’s hard, and where to start . Available at: https://www.lesswrong.com/posts/sEucKjZLbuxhQ6XdR/ai-alignment-why-it-s-hard-and-where- to-start.
