My impression from Tao’s answers is that there is a gulf between the reality of what mathematicians actually do and the benchmarks that the frontier labs tout. It doesn’t seem trivial to close that gap and create an LLM that can replace mathematicians.
LLMs are a lot dumber than you think too.
They are language prediction models, not logical deduction engines.
It says so right in the name. Expecting logic or reason or decision making is a gross mis-application.
Humans are a lot dumber than you think too.
They are self-replicating machines, not logical deduction engines.
It says so right in the name. Expecting logic or reason or decision making is a gross mis-application.
My impression from Tao’s answers is that there is a gulf between the reality of what mathematicians actually do and the benchmarks that the frontier labs tout. It doesn’t seem trivial to close that gap and create an LLM that can replace mathematicians.
Why do you sound surprised by this?
"there is a gulf between the reality of what mathematicians actually do and the benchmarks that the frontier labs tout."
A lot of peoples wealth is on the line, I wish people opened their eyes rather than believing benchmarks blindly.