Why We're getting AGI Wrong | Cog Sci and the ARC problem скачать в хорошем качестве

Why We're getting AGI Wrong | Cog Sci and the ARC problem

A look at ARC-AGI-2 (2025) results and the complete failure of Cog Sci conceptions of intelligence to shed any light on the ARC-test roadblock to AGI - this due to the capture of Cog Sci at its inception by the computer model (Information Processing) of brain/mind. Why Piaget (hugely misinterpreted) and Gibson (ignored) are critical to understanding what AGI has to achieve. #arc-agi #agi #intelligence #piaget #gibson #bergson Time Stamps 00:00 Intro - Cog Sci's capture: Gibson ignored, Piaget forced into IP 02:55 ARC-AGI-2 (just variants on ARC-1) - 2025 results 09:44 ARC-AGI-3 (2026) - dynamic - and just the beginning 11:00 Piaget - our general intelligence DEVELOPING! "Operations" = key. 15:49 Cog Sci: Imagery out! This becomes the "hard problem" 17:40 Our concept of number, operations becoming a "group" 21:37 Gibson's invariance laws - crucial for AGI, not a fit for the IP-framework 24:43 Postscript-1 - "Fluid intelligence achieved" - nope 26:37 Postscript-2 - Blindness to the problems - an index of the vacuum