New position paper argues explicit memory is central to AGI progress

A new arXiv position paper argues that explicit memory should be treated as a core requirement for progress toward AGI.

The paper contrasts today’s LLMs, which resemble implicit statistical learning, with the higher-order functions associated with hippocampal explicit memory. Those include long-term planning, metacognition, and symbolic reasoning.

Whether or not the AGI framing convinces everyone, the practical point is clear: memory architecture is becoming a central design question for advanced AI systems.