SocialAI 0.1: Towards a Benchmark to Stimulate Research on Socio-Cognitive Abilities in Deep Reinforcement Learning Agents
This work addresses the problem of building human-like social AI for researchers, but it is incremental as it introduces a benchmark without major breakthroughs.
The paper tackles the challenge of developing AI agents with socio-cognitive abilities by proposing SocialAI, a benchmark to assess social skills in deep reinforcement learning agents, and finds that a state-of-the-art approach has limits when tested on a grid-world environment.
Building embodied autonomous agents capable of participating in social interactions with humans is one of the main challenges in AI. This problem motivated many research directions on embodied language use. Current approaches focus on language as a communication tool in very simplified and non diverse social situations: the "naturalness" of language is reduced to the concept of high vocabulary size and variability. In this paper, we argue that aiming towards human-level AI requires a broader set of key social skills: 1) language use in complex and variable social contexts; 2) beyond language, complex embodied communication in multimodal settings within constantly evolving social worlds. In this work we explain how concepts from cognitive sciences could help AI to draw a roadmap towards human-like intelligence, with a focus on its social dimensions. We then study the limits of a recent SOTA Deep RL approach when tested on a first grid-world environment from the upcoming SocialAI, a benchmark to assess the social skills of Deep RL agents. Videos and code are available at https://sites.google.com/view/socialai01 .