Needs and artifcial intelligence

posted in: Research | 0

Ryan Watkins and Soheil Human just published their open access article “Needs and Artificial Intelligence” in the AI and Ethics Journal.


Throughout our history, we, Homo sapiens, have used technologies to better satisfy our needs. The relation between needs and technology is so fundamental that the US National Research Council defines the distinguishing characteristic of technology as its goal “to make modifications in the world [in order] to meet human needs” [1]. Artificial intelligence (AI) is one of the most promising emerging technologies of our time. Similar to other technologies, AI is expected by many “to meet [human] needs”. In this article, we reflect on the relationship between needs and AI, and call for the realization of needs-aware AI systems. We argue that re-thinking needs for, through, by, and with AI can be a very useful means towards the development of realistic approaches for sustainable Human-aware, Accountable, Lawful, and Ethical (HALE) AI systems. We discuss some of the most critical gaps, barriers, enablers, and drivers of co-creating future AI-based sociotechnical systems in which [human] needs are well considered and met. Finally, we provide an overview of potential challenges and considerations that should be carefully taken into account; and call for joint, immediate, and interdisciplinary efforts and collaborations to start on the path to needs-aware AI.