Bootstrap Theory of Representational Emergence: Explanatory Insufficiency as a Driver of Representation Learning and World Models
A new theory called the Bootstrap Theory of Representational Emergence (TBER) proposes that new representations in machine learning arise when existing ones become insufficient to explain observed data or transformations. This theory suggests that persistent explanatory gaps, rather than just more data or computation, drive representational innovation. TBER outlines a five-stage process from stabilized observation to provisional stabilization, applicable to various AI systems and scientific discovery. AI
IMPACT Proposes a new framework for understanding how AI systems develop more sophisticated internal representations.