I have played a little bit with OpenAI's new iteration of GPT, GPT-o1, which performs an initial reasoning step before running the LLM. It is certainly a more capable tool than previous iterations, though still struggling with the most advanced research mathematical tasks.
Here are some concrete e...
The experience seemed roughly on par with trying to advise a mediocre, but not completely incompetent, graduate student. However, this was an improvement over previous models, whose capability was closer to an actually incompetent graduate student. It may only take one or two further iterations of improved capability (and integration with other tools, such as computer algebra packages and proof assistants) until the level of "competent graduate student" is reached, at which point I could see this tool being of significant use in research level tasks.
I genuinely hate this statement. A competent grad student can solve problems. GPT cannot solve anything, as all it does is put together the shit it stole from somewhere before
This I can believe tbh. It's a very useful tool in the hands of an expert. Otherwise it's like giving a chimp a gun.
Maybe this is why I am surprised at people's hatred of ChatGPT. It's borne of misuse of a tool for experts, like newcomers struggling with a C++ compiler error.
I do agree that grad students don't exactly live in luxury, and frequently develop mental health crises. But their contributions and insight are what power their labs. Profs often have to spend so much time teaching and chasing grants that they can't do much real research. Academia overall is in a sad state.
But Tao is a superstar, and a charismatic blogger. I'd be disappointed to learn he mistreats his grad students. (I don't know if he even has any tbh)