Cognitive Revolution: GPT-6 Astra as AGI, OpenAI's Pause, and Human Control vs Technocapitalism

The Cognitive Revolution · youtube · 2026-09-12

Nathan Labenz and Prakash Narayanan host an AI:AM highlights episode centered on GPT-6 Astra's world-model capabilities and whether it qualifies as AGI, plus its "reassure and mislead" behaviors and the "no adult in the room" concern.

Key threads: OpenAI's pause and pacing agreements, safety evaluation access for outside researchers, defensive cybersecurity infrastructure built on frontier models (Amir Haghighat), children's tech without generative AI (Raffi Krikorian), US-China AI competition, and the overarching question of maintaining human control over accelerating technocapitalism. Guests include Ksenia Se, Mike Rizkalla, and Collin Hogue-Spears.

Original post →

More from AGI Musings

AGI Musings channel →