Lessons from Neuroscience for AI: How integrating Actions, Compositional Structure and Episodic Memory could enable Safe, Interpretable and Human-Like AI

TL;DR

This paper argues that integrating actions, hierarchical structure, and episodic memory from neuroscience into foundation models can improve safety, interpretability, and human-likeness in AI systems.

Contribution

It proposes a novel framework combining neuroscience-inspired components with foundation models to address current AI limitations.

Findings

Neuroscience supports the integration of actions, structure, and memory for better AI.

Adding these components can reduce hallucinations and improve grounding.

The approach enhances safety, interpretability, and energy efficiency of AI models.

Abstract

The phenomenal advances in large language models (LLMs) and other foundation models over the past few years have been based on optimizing large-scale transformer models on the surprisingly simple objective of minimizing next-token prediction loss, a form of predictive coding that is also the backbone of an increasingly popular model of brain function in neuroscience and cognitive science. However, current foundation models ignore three other important components of state-of-the-art predictive coding models: tight integration of actions with generative models, hierarchical compositional structure, and episodic memory. We propose that to achieve safe, interpretable, energy-efficient, and human-like AI, foundation models should integrate actions, at multiple scales of abstraction, with a compositional generative architecture and episodic memory. We present recent evidence from neuroscience and cognitive science on the importance of each of these components. We describe how the addition of these missing components to foundation models could help address some of their current deficiencies: hallucinations and superficial understanding of concepts due to lack of grounding, a missing sense of agency/responsibility due to lack of control, threats to safety and trustworthiness due to lack of interpretability, and energy inefficiency. We compare our proposal to current trends, such as adding chain-of-thought (CoT) reasoning and retrieval-augmented generation (RAG) to foundation models, and discuss new ways of augmenting these models with brain-inspired components. We conclude by arguing that a rekindling of the historically fruitful exchange of ideas between brain science and AI will help pave the way towards safe and interpretable human-centered AI.