Artificial Expert Intelligence through PAC-reasoning
This addresses the need for more reliable and precise AI reasoning in domain-specific expert tasks, though it appears incremental as it builds on existing PAC theory and System 1/2 concepts.
The paper tackles the problem of AI systems lacking adaptability and precision in novel problem-solving by introducing Artificial Expert Intelligence (AEI) with a PAC-reasoning framework, which provides robust theoretical guarantees for decomposing complex problems and controlling reasoning precision, establishing a foundation for error-bounded, inference-time learning.
Artificial Expert Intelligence (AEI) seeks to transcend the limitations of both Artificial General Intelligence (AGI) and narrow AI by integrating domain-specific expertise with critical, precise reasoning capabilities akin to those of top human experts. Existing AI systems often excel at predefined tasks but struggle with adaptability and precision in novel problem-solving. To overcome this, AEI introduces a framework for ``Probably Approximately Correct (PAC) Reasoning". This paradigm provides robust theoretical guarantees for reliably decomposing complex problems, with a practical mechanism for controlling reasoning precision. In reference to the division of human thought into System 1 for intuitive thinking and System 2 for reflective reasoning~\citep{tversky1974judgment}, we refer to this new type of reasoning as System 3 for precise reasoning, inspired by the rigor of the scientific method. AEI thus establishes a foundation for error-bounded, inference-time learning.