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

How is ChatGPT's behavior changing over time?

arxiv.org

Post reply on HN
Print view View original

111–120 of 187 posts

Re: How is ChatGPT's behavior changing over time?

#111

Post by wizofaus »

SanderNL wrote: post #98
xp84 wrote: post #43

Earlier quoted context omitted.

Seriously. GPT doing math is like using a 737 to drive around on the ground, or if you had the phone number of a prominent astrophysicist and you call him to do long division for you. Wtf is the point. We have computer things to do every math problem. It’s a waste of energy to use LLMs for it in my opinion.

It’s not about the results, it’s about its ability to “reason”. Math is about as close to pure reasoning we get so I don’t get the pessimism. If it is bad at math and can’t be taught, then you have a fundamental problem. It’s a matter of time before this limit gets hit in other domains.

This seems fair enough to me. That ChatGPT currently struggles with certain types of maths problems points to reasonably fundamental shortcomings in what otherwise appears to have the beginnings of a general purpose reasoning engine (whether you consider it AGI or not), and I'm willing to bet extraordinarily clever minds are working hard on trying to address those shortcomings.

Re: How is ChatGPT's behavior changing over time?

#112

Post by marci »

Closi wrote: post #104
xp84 wrote: post #43

Earlier quoted context omitted.

Seriously. GPT doing math is like using a 737 to drive around on the ground, or if you had the phone number of a prominent astrophysicist and you call him to do long division for you. Wtf is the point. We have computer things to do every math problem. It’s a waste of energy to use LLMs for it in my opinion.

> Seriously. GPT doing math is like using a 737 to drive around on the ground, or if you had the phone number of a prominent astrophysicist and you call him to do long division for you. Wtf is the point. We have computer things to do every math problem. It’s a waste of energy to use LLMs for it in my opinion. Using GPT to do maths is probably like using a 737 to drive around on the ground. Teaching GPT to do maths mi…

Apparently, models finetuned for coding are better at logical inquiries than those that are not.

Re: How is ChatGPT's behavior changing over time?

#113

Post by falcor84 »

dudeinhawaii wrote: post #6

I think we should stop trying to quiz LLMs on mathematics, something for which they are explicitly not designed to do with their tokenized view of the world. Ask GPT-4 to use its Wolfram plugin and it returns the answers quickly and correctly. Second, I think the code generation bit of this paper is blown out of proportion. The code can't be immediately injected into a codebase due to a formatting change (triple quot…

I would argue that ChatGPT is actually quite good at "mathematics", in the sense of helping me formulate a problem in an appropriate mathematical structure, and coming up with a good sequence of steps to solve / simplify it. Not perfect by any means, but not bad at all.

What it is bad at is actually performing the steps accurately, but as others mentioned, that's where Wolfram and/or a code interpreter would come in.

Re: How is ChatGPT's behavior changing over time?

#114

Post by falcor84 »

muzani wrote: post #96
deelly wrote: post #72

Earlier quoted context omitted.

No, we definitely should continue to quiz LLMs on mathematics and absolutely any other topics. Otherwise how do we know and understand limitation of the system?

We should also test its capabilities on cooking steak, flying rockets, and making love. Only then will we know if AI can be superior to humans on all things.

I know you're being facetious, but I'd absolutely be in favor of having AI benchmarks for any and all of these.

Re: How is ChatGPT's behavior changing over time?

#115

Post by WithinReason »

Traubenfuchs wrote: post #103
josefx wrote: post #99

Earlier quoted context omitted.

wouldn't it be less time to just type one yourself? Like open notepad, hit the keys like a deranged monkey, select the first 64 characters, done?

That is very much not random, but generally enough for 99% of all use cases.

It would be more random than an LLM's output

Re: How is ChatGPT's behavior changing over time?

#116

Post by ineedasername »

blobbers wrote: post #79
ineedasername wrote: post #33

Earlier quoted context omitted.

>I read the paper, and dumbed it down to dogs and cats. I could read the Pricicipia, or anything, and dumb it down to dogs and cats, but that reduction to the dumb would be a failing all my own. As for the test-- a classification task benchmarked with 500 examples is a fairly decent test of a system's capabilities, be it LLM or a traditional machine learning model. And again, this was only one of a variety of tasks.…

Well, for starters, you take a model with 100 trillion parameters, and then you test it with... 500 examples. WAIT did you say hundred? 500 hundred? Not 500 hundred thousand right? Okay okay hold on back to the drawing board. ...dumbed down to cats and dogs.

This is simply not the way such models work and your jump from the LLM’s parameter count in its own training to testing its capabilities in a classification task is a non sequitur.

…Principia and failings

Re: How is ChatGPT's behavior changing over time?

#117

Post by jstummbillig »

xp84 wrote: post #43
dudeinhawaii wrote: post #6

I think we should stop trying to quiz LLMs on mathematics, something for which they are explicitly not designed to do with their tokenized view of the world. Ask GPT-4 to use its Wolfram plugin and it returns the answers quickly and correctly. Second, I think the code generation bit of this paper is blown out of proportion. The code can't be immediately injected into a codebase due to a formatting change (triple quot…

Seriously. GPT doing math is like using a 737 to drive around on the ground, or if you had the phone number of a prominent astrophysicist and you call him to do long division for you. Wtf is the point. We have computer things to do every math problem. It’s a waste of energy to use LLMs for it in my opinion.

"We" are never going to stop trying to use LLMs for math. They are obviously mimicking a smart person. What do you do with smart persons? You query them with bad and lazy questions (often without being too honest about that to yourself), and hope/expect helpful answers.

Because that is how it works. The limiting factor so far has been the smart persons time and patience. Now, no longer.

People moderating their LLMs usage is never happening, from here on out until the end of civilisation. Any LLM service that is designing for that is done. You need to make lazy questions efficient. People do not care about how complicated your sql query is and they will never care. People will not give up on energy, meat, cars, as long as they feel they are giving something up.

People will never think twice to not make your LLM think twice.

If it seems useful and convenient, people will use it. If it's not giving good answers to lazy questions out of the box, they will go to the thing that does.

Re: How is ChatGPT's behavior changing over time?

#118

Post by amelius »

How is Copilot's behavior changing over time?

And has anyone successfully used Copilot to do simple but tedious things like generate library bindings for various languages? I suspect that Copilot would be good at this kind of thing because it is a relatively simple task and there is a lot of training-code available.

Re: How is ChatGPT's behavior changing over time?

#119

Post by remus »

xp84 wrote: post #43
dudeinhawaii wrote: post #6

I think we should stop trying to quiz LLMs on mathematics, something for which they are explicitly not designed to do with their tokenized view of the world. Ask GPT-4 to use its Wolfram plugin and it returns the answers quickly and correctly. Second, I think the code generation bit of this paper is blown out of proportion. The code can't be immediately injected into a codebase due to a formatting change (triple quot…

Seriously. GPT doing math is like using a 737 to drive around on the ground, or if you had the phone number of a prominent astrophysicist and you call him to do long division for you. Wtf is the point. We have computer things to do every math problem. It’s a waste of energy to use LLMs for it in my opinion.

> GPT doing math is like using a 737 to drive around on the ground, or if you had the phone number of a prominent astrophysicist and you call him to do long division for you.

I don't think this is a great analogy. if your 737 couldn't drive on the ground and your astrophysicist couldn't answer basic maths questions I wouldn't want to fly in that plane or put much faith in the astrophysicists answers to more complex questions.

Maybe maths is not a particular strength of LLMs, but asking questions where it is easy to judge the factual accuracy of the responses seems a pretty reasonable test to be running.

Re: How is ChatGPT's behavior changing over time?

#120

Post by Kubuxu »

TeMPOraL wrote: post #89
dudeinhawaii wrote: post #6

I think we should stop trying to quiz LLMs on mathematics, something for which they are explicitly not designed to do with their tokenized view of the world. Ask GPT-4 to use its Wolfram plugin and it returns the answers quickly and correctly. Second, I think the code generation bit of this paper is blown out of proportion. The code can't be immediately injected into a codebase due to a formatting change (triple quot…

In terms of evaluating LLMs, I'd argue quizzing them on maths is still better than the other thing people keep doing - quizzing them on facts and self-contradicting scenarios, hoping to get them to only either recall information perfectly, or answer with "I don't know". I'm not an AI/ML scientist, so I may be way off mark here, but everything I've read so far, and all my experience playing with GPT-3.5 and GPT-4, tel…

I couldn’t agree with you more. Today when we are talk with LLM it’s like we (sorry for anthropomorphism) interact with a naked mind and we rely on pure immediate recall and it’s train of thought. I wonder whether allowing LLMs to perform inner dialogue (as it was done with the through/action pattern), to stage information for latter, and then based on that making it form response will be the next step.

It will be much more compute intensive, as each response will probably require multiple context windows and distillations.

Post reply on HN