Why AI systems don't learn and what to do about it: Lessons on autonomous learning from cognitive science

Researchers propose a new AI learning architecture inspired by human and animal cognition that integrates observational learning and active behavior learning. The framework includes a meta-control system that switches between learning modes, addressing current limitations in autonomous AI learning.

• →Current AI systems have significant limitations in achieving autonomous learning compared to biological systems.
• →The proposed architecture combines System A (observational learning) and System B (active behavior learning).
• →A meta-control System M manages flexible switching between different learning modes.
• →The framework draws inspiration from how organisms adapt across evolutionary and developmental timescales.
• →This approach could advance AI towards more autonomous and adaptive learning capabilities.