Lots of questions around what was really shown in the FrontierMath benchmark, too. We need transparency, and instead got hype.
@garymarcus
-
O3 unreliable across domains, better AI technology will emerge
By
–
Aside from details around ARC & other benchmarks, o3 won’t turn out to be reliable across domains, and it will in hindsight seem wildly inefficient. MUCH better technology will emerge, after the LLM hype subsides. (LLMs may still exist, but will be just one tool in a larger
-
AlphaFold: A Greater Contribution Than ChatGPT
By
–
and, not unrelatedly AlphaFold is IMHO a much greater contribution to the world than ChatGPT
-
o3 Not Close to AGI Says Expert in 2026
By
–
Mark my words: when AGI does actually come, perhaps 10 or 20 years from now, people will laugh at the idea that o3 was close to AGI.
-
OpenAI Pays Workers $100/Hour for Data Labeling Tasks
By
–
Retweeting, roughly two years later. This was exactly right, except that now OpenAI is paying some people $100/hour to do it.
-
Questioning o3’s True Progress Toward AGI Development
By
–
But also, given that we agree that o3’s prowess was overreported, and given that we didn’t see any really clean test of OOD challenges, and given that we don’t know what was in the data augmentation, why are you confident that o3 is an important step towards AGI? Another
-
Methodology Matters in AI Research Presentations
By
–
So do I! Just not soon. What is impressive is that you are more on the “soon” side than I am — but reached similar conclusions about the shortcomings in the presentations. I was heavily attacked for making very similar points. But like you I think methodology matters.
-
Science and Technology: From Pasteurization to Modern Innovation
By
–
It is also the science that gave us cell phones and the Internet, pasteurization, airplanes, rockets, etc
-
LeCun Misrepresents My Deep Learning Position on Threads
By
–
. @yLeCun continues to lie about my views on Threads, saying e.g. that I have argued that “Deep Learning [will not] remain the foundation of present and future AI systems” when I never said any such thing. What I have *actually* said over and over is deep learning is possibly
-
Why Scientists Should Fear AI Development Risks
By
–
Everyone should be terrified. Not just scientists, but anyone who depends on science and well-trained scientists: tech companies, pharma companies, ordinary citizens, etc.