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

71–80 of 527 posts

Re: Terence Tao on O1

#71

Post by wenc »

threeseed wrote: post #45
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…

> HN, and the internet in general, have become just an ocean of reactionary sandbagging and blather about how "useless" LLMs are. This is cult like behaviour that reminds me so much of the crypto space. I don't understand why people are not allowed to be critical of a technology or not find it useful. And if they are they are somehow ignorant, over-reacting or deficient in some way.

Also, I'm deeply skeptical of crypto too due to its present scamminess, but I am keeping an open mind that there is a future in which crypto -- once it gets over this phase of get-rich-quick schemers -- will be seen as just another asset class.

I read somewhere that historically bonds in their early days were also associated with scamminess but today they're just a vanilla asset.

Re: Terence Tao on O1

#72

Post by threeseed »

wenc wrote: post #65
threeseed wrote: post #45

Earlier quoted context omitted.

> HN, and the internet in general, have become just an ocean of reactionary sandbagging and blather about how "useless" LLMs are. This is cult like behaviour that reminds me so much of the crypto space. I don't understand why people are not allowed to be critical of a technology or not find it useful. And if they are they are somehow ignorant, over-reacting or deficient in some way.

I think it's perfectly ok to be critical of technology as long as one is thoughtful rather than dismissive. There is a lot of hype right now and pushing back against it is the right thing to do. I'm more reacting against simplistic and categorical pronouncements of straight up "uselessness," which to me seems un-curious and deeply cynical, especially since it is evidentially untrue in many domains (though it is true…

This entire comment can be summarised as: everyone who doesn't think like me is wrong.

Not everyone is interested in seeing the world through the hopes and dreams of e/acc types and would prefer to see it as it is today.

LLMs are a technology. Nothing more. It can be as amazing or useless as anyone likes.

Re: Terence Tao on O1

#74

Post by apsurd »

vasco wrote: post #69
benterix wrote: post #38

Earlier quoted context omitted.

> I've found that I haven't written a line of code in weeks Which is great until your next job interview. Really, it's tempting in the short run but I made a conscious decision to do certain tasks manually only so that I don't lose my basic skills.

ChatGPT voice interface plugged into the audio stream, with the prompt: - I need you to assist me during a programming interview, you will be listening to two people, the interviewer and me. When the interviewer asks a question, I'd like you to feed me lines that seem realistic for an interview where I'm nervous, don't give me a full blown answer right away. Be very succinct. If I think you misunderstood something, I…

This made me laugh. I can't deny it isn't already happening. But wow people work so hard to avoid working hard.

Re: Terence Tao on O1

#75

Post by jdiez17 »

eigenvalue wrote: post #64

The o1 model is really remarkable. I was able to get very significant speedups to my already highly optimized Rust code in my fast vector similarity project, all verified with careful benchmarking and validation of correctness. Not only that, it also helped me reimagine and conceptualize a new measure of statistical dependency based on Jensen-Shannon divergence that works very well. And it came up with a super fast i…

> 1,337 additions

cough

Re: Terence Tao on O1

#76

Post by kldnav »

Tao and Aaronson are optimistic about LLMs. What are they telling their students? That math and science degrees will soon have the same value as a degree in medieval dance theory?

If they are overly optimistic, perhaps it would be good to hear the opinions of Wiles and Perelman.

Re: Terence Tao on O1

#77

Post by ein0p »

Idk I think the fact that it needs “hints” and “prodding” is a good thing, myself. Otherwise we wouldn’t need humans to get those answers, would we. I want it to augment humans, not replace them.

Re: Terence Tao on O1

#79

Post by adt »

sgt101 wrote: post #36

Is there a list of discoveries or siginficant works/constructions made by people collaborating with LLM's? I mean as opposed to specific deep networks like Alphafold or Graphcast?

I'd like to see that, too.

I have a related list of GPT accomplishments here: https://docs.google.com/spreadsheets/d/1kc262HZSMAWI6FVsh0zJ...

Re: Terence Tao on O1

#80

Post by jamesmotherway »

benterix wrote: post #38
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 Which is great until your next job interview. Really, it's tempting in the short run but I made a conscious decision to do certain tasks manually only so that I don't lose my basic skills.

Not everyone is doing coding interviews. Some might struggle with a particular language due to lack of muscle memory, but can dictate the logic in pseudocode and can avoid pitfalls inferred from past experience. This sort of workflow is compatible with LLMs, assuming a sufficient background (otherwise one can't recognize when the output diverges from your intent).

I personally treat the LLM as a rubber duck. Often I reject its output. In other cases, I can accept it and refactor it into something even better. The name of the game is augmentation.

Post reply on HN