This paper says today's content AIs are great at pretty pictures and videos but often miss what people actually want, creating a big Intent-Execution Gap.
This paper explains how to turn large language models (LLMs) from quiet students that only answer questions into active agents that can plan, act, and learn over time.