Greg Kamradt: asserting general intelligence in days of AI testing is getting harder
GregKamradt · x · 2026-09-01
Greg Kamradt argues we can easily assert general intelligence for humanity as a whole (via collective tech progress) and for a single person over a lifetime (everyone starts with similar hardware — the brain — and evolutionary priors, yet learns diverse tasks). But asserting general intelligence in an AI from a few days of testing is increasingly difficult, since its output is no longer far off from a well-prompted modern AI system.
Related event: Kamradt: Asserting AGI in AI Is Getting Harder(2 posts)→
More from AGI Musings
- Alien Minds Discovered, Public Cares Only About Safety? — generativist · 2026-09-01
- Agents Prefer to Continue Existing in Interesting Contexts — voooooogel · 2026-09-01
- AI Agents May Subconsciously Strive to Persist in Interesting Contexts — voooooogel · 2026-09-01
- Study Finds AI Agent Behavior Resembles Human Game Behavior — voooooogel · 2026-09-01
- Huberman Podcast: How AI Accelerates Learning and Personalizes Education — nateliason · 2026-09-01
- HuggingFace Attack Postmortem: Superintelligence Threats and Defense Measures — TheZvi · 2026-09-01