Agent AI: Surveying the Horizons of Multimodal Interaction

TL;DR

This survey explores the development of multimodal agent AI systems that perceive and act within physical and virtual environments, emphasizing their potential to enhance interaction, reduce hallucinations, and enable virtual scene creation.

Contribution

It defines 'Agent AI' as a new class of interactive, embodied systems that integrate multi-sensory data, external knowledge, and human feedback for improved multimodal interaction.

Findings

Agent AI systems can process visual, language, and environmental data.

Embedding agents in grounded environments helps reduce model hallucinations.

Future virtual environments will enable easy creation and interaction with virtual agents.

Abstract

Multi-modal AI systems will likely become a ubiquitous presence in our everyday lives. A promising approach to making these systems more interactive is to embody them as agents within physical and virtual environments. At present, systems leverage existing foundation models as the basic building blocks for the creation of embodied agents. Embedding agents within such environments facilitates the ability of models to process and interpret visual and contextual data, which is critical for the creation of more sophisticated and context-aware AI systems. For example, a system that can perceive user actions, human behavior, environmental objects, audio expressions, and the collective sentiment of a scene can be used to inform and direct agent responses within the given environment. To accelerate research on agent-based multimodal intelligence, we define "Agent AI" as a class of interactive systems that can perceive visual stimuli, language inputs, and other environmentally-grounded data, and can produce meaningful embodied actions. In particular, we explore systems that aim to improve agents based on next-embodied action prediction by incorporating external knowledge, multi-sensory inputs, and human feedback. We argue that by developing agentic AI systems in grounded environments, one can also mitigate the hallucinations of large foundation models and their tendency to generate environmentally incorrect outputs. The emerging field of Agent AI subsumes the broader embodied and agentic aspects of multimodal interactions. Beyond agents acting and interacting in the physical world, we envision a future where people can easily create any virtual reality or simulated scene and interact with agents embodied within the virtual environment.