Abstract
This paper proposes use of computational models of argumentation based dialogue for enhancing the quality and scope ('scaffolding') of both human and artificial agent reasoning. In support of this proposal I draw on work in cognitive psychology that justifies such a role for human reasoning. I also refer to recent concerns about the potential dangers of artificial intelligence (AI), and the consequent need to ensure that AI actions are aligned with human values. I will advocate that argumentation based models of dialogue can contribute to value alignment by enabling joint human and AI reasoning that may indeed be better purposed to resolve challenging ethical issues. This paper also reviews research in formals models of argumentation based reasoning and dialogue that will underpin applications for scaffolding human and artificial agent reasoning.
| Original language | English |
|---|---|
| Pages (from-to) | 58-71 |
| Number of pages | 14 |
| Journal | CEUR Workshop Proceedings |
| Volume | 2090 |
| Publication status | Published - 1 Jan 2018 |
| Event | 5th International Workshop on Artificial Intelligence and Cognition, AIC 2017 - Larnaca, Cyprus Duration: 1 Nov 2017 → 3 Nov 2017 |