knotthebest

3 ms·

1.
by knotthebest 2y ago
Do note that Terry has access to the full o1. o1-preview is, well, a preview.
2.
by knotthebest 2y ago
They are, though. I spoke to an author just yesterday. They did mostly use brute-force.
3.
by knotthebest 2y ago
Still, very marginal improvements. There is also Amdahl’s Law which limits the speed gains of parallelisation.
4.
by knotthebest 2y ago
Skill issue tbh
5.
by knotthebest 2y ago
You’re not understanding the tests. They were designed for humans . They were also designed to be taken once. LLMs have been trained on these tests numerous times. LLMs are also not humans. We can’t conceivably compare the two
6.
by knotthebest 2y ago
Well, they were trained on those tests. What LLMs literally do is map inputs to outputs. That’s why they can’t do much outside of training data.
7.
by knotthebest 2y ago
I just don’t think Ray Kurzweil understands these systems well enough. His views aren’t of much substance anymore.
8.
by knotthebest 2y ago
Nothing particularly interesting to my knowledge. You should read this though, some people at Symbolica AI worked on it too: https://www.google.com/url?q=https://arxiv.org/abs/2402.1533...
9.
by knotthebest 2y ago
I’m not sure how this would solve the hard problem. Could you please elaborate on that?
10.
by knotthebest 3y ago
My guess too
11.
by knotthebest 3y ago
I think I may be missing something. Why would you be interested in the automation of theorem production? Wouldn’t this make mathematicians obsolete? How far away do you think we are from that? I ask as a newbie in math; math is a passion of
12.
by knotthebest 3y ago
At which point would mathematicians become obsolete? Something like this seems like it could automate a lot of mathematics research, no?
13.
by knotthebest 3y ago
I’m not so sure of this. The fear generally isn’t that the model itself is dangerous, more so that how it is used by someone might be.
14.
by knotthebest 3y ago
Eh
15.
by knotthebest 3y ago
Generally just curious about what others a bit more knowledgeable in this field than me might think.
16.
Terence Tao, "Machine Assisted Proof" [video](youtube.com)
2 points by knotthebest 3y ago|1 comments