Measuring progress toward AGI: A cognitive framework

A new paper, "Measuring Progress Toward AGI: A Cognitive Taxonomy," presents a framework that draws on cognitive science to assess AI systems' general intelligence and to provide empirical tools for tracking progress toward AGI. The taxonomy identifies ten cognitive abilities believed to be important for general intelligence: perception, generation, attention, learning, memory, reasoning, metacognition, executive functions, problem solving and social cognition.

To benchmark systems against human capabilities, the paper proposes a three-stage evaluation protocol: evaluate AI across a broad suite of cognitive tasks using held-out test sets, collect human baselines from a demographically representative sample of adults, and map each AI system's performance relative to the distribution of human performance for each ability.

agi, cognitive taxonomy, general intelligence, cognitive abilities, evaluation protocol, human baselines, held-out tests, benchmarking, metacognition, executive functions