Model Releases
Jensen Huang says that AGI has been achieved. I strongly disagree. GPT-6 performs really well on benchmarks, including ARC-AGI-3. But benchm…
Jensen Huang says that AGI has been achieved. I strongly disagree. GPT-6 performs really well on benchmarks, including ARC-AGI-3. But benchmark performance is not the same as general intelligence. GPT
Jensen Huang says that AGI has been achieved. I strongly disagree. GPT-6 performs really well on benchmarks, including ARC-AGI-3. But benchmark performance is not the same as general intelligence. GPT-6 is great at many tasks with fixed goals. It can even solve extremely difficult mathematical problems. But general intelligence is not only solving difficult problems that no human can solve. It is also solving simple but unfamiliar problems that every ten-year-old kid can solve. AGI is not just about solving closed-ended problems. It is about navigating an open-ended world. * Full piece in the first reply. P.S. A new paper on adaptive intelligence will come out in about a month. Stay tuned.
Related
- doomers and AI fans love hearing “it’s over” - but @scaling01 really should have noted that the below refers to a domain specific benchmark …
- so much for the “general” part of general intelligence
- Just because someone puts the word AGI in their benchmark doesn't make it even remotely a reflection of what the general public understands …
- even within math, frontier models are far from being general intelligences. and we still need people to assess which of their output are tru…
- We stress tested many frontier AI models for multimodal medical reasoning (including GPT-5, Claude 3.5, Gemini 2.5 Pro). They’re not ready. …
Source: Gary Marcus (X) | 2026-09-07