Modern AI systems are increasingly evaluated for their reasoning, coding, theorem proving, tool usage, and long-term research capabilities. These powerful capabilities share a structural limitation: the representational frame within which the model operates, including its conceptual vocabulary, the space of admissible solutions it can search, and the criteria for success, is typically fixed and supplied in advance.
This paper argues that building stronger intelligent systems capable of open-ended innovation requires additional classes of operations: the creation, stabilization, and reuse of new representational primitives, which alter the search space rather than simply searching within it.
We characterize the distance between current AI systems and genuinely open-ended intelligence through two gaps. The first is the vocabulary gap, the difficulty of inventing and stabilizing new representational primitives rather than merely recombining existing ones. The second is the verifier gap, the difficulty of judging the value of a new primitive when its full payoff may only be visible after future reuse.
We interpret both gaps through a unified framework of intelligence as cognitive discrepancy reduction. By viewing intelligent behaviors as a sequence of cognitive transformations, we distinguish intra-space transformations that operate within a fixed representational frame from generative transformations that may modify the frame itself.
Based on this, we propose a ladder of innovation autonomy and outline several directions for advancing open-ended AI, including objectives that reward useful representational change, persistent memory architectures for invented primitives, and adaptive verification mechanisms capable of evolving alongside the representations they evaluate.
Blogger's Review: This paper delves into the limitations of AI systems in open-ended innovation, emphasizing how introducing new representational primitives can propel intelligence development. The analysis of vocabulary and verifier gaps provides crucial insights for future AI research, highlighting the importance of dynamic adaptability and innovative capacity.