Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Terence Tao on O1

mathstodon.xyz

Post reply on HN
Print view View original

281–290 of 527 posts

Re: Terence Tao on O1

#281

Post by Kiro »

skydhash wrote: post #68
WhatIsDukkha wrote: post #26

Earlier quoted context omitted.

I entirely agree about their utility. HN, and the internet in general, have become just an ocean of reactionary sandbagging and blather about how "useless" LLMs are. Meanwhile, in the real world, I've found that I haven't written a line of code in weeks. Just paragraphs of text that specify what I want and then guidance through and around pitfalls in a simple iterative loop of useful working code. It's entirely a lea…

> Much Much more productive world by just knuckling down and learning how to do the work. The fact everyone that say they've become more productive with LLMs won't say how exactly. I can talk about how VIM have make it more enjoyable to edit code (keybinding and motions), how Emacs is a good environment around text tooling (lisp machine), how I use technical books to further my learning (so many great books out here)…

You clearly have made up your mind that it can't be right but to me it's like arguing against breathing. There are no uncertainties or misunderstandings here. The productivity gains are real and the code produced is more robust. Not in theory, but in practice. This is a fact for me and you trying to convince me otherwise is just silly when I have the result right in front of me. It's also not just boilerplate. It's all code.

Re: Terence Tao on O1

#282

Post by Workaccount2 »

skydhash wrote: post #68
WhatIsDukkha wrote: post #26

Earlier quoted context omitted.

I entirely agree about their utility. HN, and the internet in general, have become just an ocean of reactionary sandbagging and blather about how "useless" LLMs are. Meanwhile, in the real world, I've found that I haven't written a line of code in weeks. Just paragraphs of text that specify what I want and then guidance through and around pitfalls in a simple iterative loop of useful working code. It's entirely a lea…

> Much Much more productive world by just knuckling down and learning how to do the work. The fact everyone that say they've become more productive with LLMs won't say how exactly. I can talk about how VIM have make it more enjoyable to edit code (keybinding and motions), how Emacs is a good environment around text tooling (lisp machine), how I use technical books to further my learning (so many great books out here)…

I have used chatGPT to write test systems for our (physical) products. I have a pretty decent understanding of how code/programs works structurally, I just don't know the syntax/language (Python in this case).

So I can translate things like

"Create an array, then query this instrument for xyz measurements, then store those measurements in the array. Then store that array in the .csv file we created before"

It works fantastic and saved us from outsourcing.

Re: Terence Tao on O1

#283

Post by acedTrex »

mewpmewp2 wrote: post #279
acedTrex wrote: post #269

Earlier quoted context omitted.

I have it, tried it for a while. I have it turned mostly off new except for rare boilerplate heavy cases. It kept generating annoyingly wrong code. Things with subtly wrong misleading names, missing edge cases, ignoring immediate same file context etc. I found that it slowed me down so i turned it off.

Which language?

golang and python mainly.

For rust it failed spectacularly. So bad that its not worth discussing lol

Re: Terence Tao on O1

#284

Post by sterlind »

wenc wrote: post #5

Once GPT is tuned more heavily on Lean (proof assistant) -- the way it is on Python -- I expect its usefulness for research level math to increase. I work in a field related to operations research (OR), and ChatGPT 4o has ingested enough of the OR literature that it's able to spit out very useful Mixed Integer Programming (MIP) formulations for many "problem shapes". For instance, I can give it a logic problem like "…

I also do OR-adjacent work, but I've had much less luck using 4o for formulating MIPs. It tends to deliver correct-looking answers with handwavy explanations of the math, but the equations don't work and the reasoning doesn't add up.

It's a strange experience, like taking a math class where the proofs are weird and none of the lessons click for you, and you start feeling stupid, only to learn your professor is an escaped dementia patient and it was gobbledygook to begin with.

I had a similar experience yesterday using o1 to see if a simple path exists through s to t through v using max flow. It gave me a very convincing-looking algorithm that was fundamentally broken. My working solution used some techniques from its failed attempt, but even after repeated hints it failed to figure out a working answer (it stubbornly kept finding s->t flows, rather than realizing v->{s,t} was the key.)

It's also extremely mentally fatiguing to check its reasoning. I almost suspect that RLHF has selected for obfuscating its reasoning, since obviously-wrong answers are easier to detect and penalize than subtly-wrong answers.

Re: Terence Tao on O1

#285

Post by TrackerFF »

fnordpiglet wrote: post #149

Rewind your mind to 2019 and imagine reading a post that said “The experience seemed roughly on par with trying to advise a mediocre, but not completely incompetent, graduate student.” With regard to interacting with the equivalent of Alexa. That’s a remarkable difference in 5 years.

Even more amazing, there plenty - PLENTY - of posters here that routinely either completely shit on LLMs, or casually dismiss them as "hype", "useless", and what have you.

I've been saying this for quite some time now, but some people are in for a very rude awakening when the SOTA models 5-10 years from now are able to completely replace senior devs and engineers.

Better buckle up, and start diversifying your skills.

Re: Terence Tao on O1

#286

Post by Davidzheng »

asdasjhG wrote: post #197
Davidzheng wrote: post #178

Earlier quoted context omitted.

What does this mean? Of course math AI will take over top research in next ten years but usefulness to society has never been a goal of pure mathematics. I don't know if you understand the motivation for studying pure math. Personally I think it will be mostly good for research math

The "value of a degree" means the employment prospects for the degree holder. Which is going to zero if the optimistic predictions are correct, so the optimistic professors should warn their students. I understand the motivation for pure math quite well. It is about beauty, understanding things and discovering things for oneself. If computers do the work, the discovery part is gone and pure math is ruined. For the no…

do you also value your personal relationships based on employment prospects?

Re: Terence Tao on O1

#287

Post by zeroonetwothree »

TrackerFF wrote: post #285
fnordpiglet wrote: post #149

Rewind your mind to 2019 and imagine reading a post that said “The experience seemed roughly on par with trying to advise a mediocre, but not completely incompetent, graduate student.” With regard to interacting with the equivalent of Alexa. That’s a remarkable difference in 5 years.

Even more amazing, there plenty - PLENTY - of posters here that routinely either completely shit on LLMs, or casually dismiss them as "hype", "useless", and what have you. I've been saying this for quite some time now, but some people are in for a very rude awakening when the SOTA models 5-10 years from now are able to completely replace senior devs and engineers. Better buckle up, and start diversifying your skills.

Not sure how this post is evidence of AIs replacing senior devs.

Re: Terence Tao on O1

#288

Post by dartos »

mewpmewp2 wrote: post #261
dartos wrote: post #254

Earlier quoted context omitted.

> At this point, such sentiments feel either willfully ignorant, or said in bad faith. I feel exactly the same, but in the opposite direction. As someone who’s been programming for 17 years and working professionally for 10, I’m unable to get any huge productivity boosts from AI tools. They’re better than Google+stack overflow for asking random questions, but in a specific context and they’re good for repetitive, but…

Most gains are from using Copilot, do you use that?

I use it for those refactors I mentioned in my comment.

It’s autocomplete++, except without knowledge of the rest of my codebase.

Re: Terence Tao on O1

#289

Post by dartos »

acedTrex wrote: post #269
mewpmewp2 wrote: post #261

Earlier quoted context omitted.

Most gains are from using Copilot, do you use that?

I have it, tried it for a while. I have it turned mostly off new except for rare boilerplate heavy cases. It kept generating annoyingly wrong code. Things with subtly wrong misleading names, missing edge cases, ignoring immediate same file context etc. I found that it slowed me down so i turned it off.

This is my exact experience as well.

Re: Terence Tao on O1

#290

Post by Kiro »

wokwokwok wrote: post #256
WhatIsDukkha wrote: post #26

Earlier quoted context omitted.

I entirely agree about their utility. HN, and the internet in general, have become just an ocean of reactionary sandbagging and blather about how "useless" LLMs are. Meanwhile, in the real world, I've found that I haven't written a line of code in weeks. Just paragraphs of text that specify what I want and then guidance through and around pitfalls in a simple iterative loop of useful working code. It's entirely a lea…

> I've found that I haven't written a line of code in weeks Please post a video of your workflow. It’s incredibly valuable for people to see this in action, otherwise they, quite legitimately, will simply think this is not true.

Who cares what they think? In fact, the fewer who uses this the better for the ones that do. It's not in my self-interest to convert anyone and I obviously don't need to convince myself when I have the result right in front of me. Whether you believe it or not does not make me less productive.
Post reply on HN