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

GPT-4

openai.com

Post reply on HN
Print view View original

511–520 of 1001 posts

Re: GPT-4

#511

Post by TallGuyShort »

lastangryman wrote: post #467

Genuinely surprised by the positive reaction about how exciting this all is. You ever had to phone a large business to try and sort something out, like maybe a banking error, and been stuck going through some nonsense voice recognition menu tree that doesn't work? Well imagine chat GPT with a real time voice and maybe a fake, photorealistic 3D avatar and having to speak to that anytime you want to speak to a doctor,…

I would hardly consider my previous experiences dealing with doctors, tax administrators, mortgage companies, or recruiters to be anything close to good models of what human interaction should look like. In fact all of these people might be close to the top of the list of the most unpleasant interactions I've ever had. I'm at least willing to see what it looks like when they adopt AI for communication.

Re: GPT-4

#513

Post by amai »

I would love if GPT-4 would be connected to github and starts to solve all open bugs there. Could this be the future: Pull requests from GPT-4 automatically solving real issues/problems in your code?

Re: GPT-4

#515

Post by colordrops »

lastangryman wrote: post #467

Genuinely surprised by the positive reaction about how exciting this all is. You ever had to phone a large business to try and sort something out, like maybe a banking error, and been stuck going through some nonsense voice recognition menu tree that doesn't work? Well imagine chat GPT with a real time voice and maybe a fake, photorealistic 3D avatar and having to speak to that anytime you want to speak to a doctor,…

No matter how accurate, GPT can't fake domain names. You can still determine where information came from. So, as it has always been, you decide which sources you trust. You get information from a particular website, and it turns out to be true and works out for your needs, so you trust it in the future. Whether the information on that site is from humans or from AI is not material.

The situation is not different from now. Humans have been faking information from the beginning of time. The only difference is scale. Perhaps this will be a good thing, as fakery was limited enough to slip through the cracks, but now everyone will be forced to maintain a critical eye, and verify sources and provenance.

Re: GPT-4

#516

Post by eternalban »

Greg Brockman just tldr'd the whole thing in his live deeloper demo of GPT-4:

~ "GPT-4. It's not perfect, but neither are you"

Re: GPT-4

#517

Post by kristianp »

ml_basics wrote: post #136

From the paper: > Given both the competitive landscape and the safety implications of large-scale models like GPT-4, this report contains no further details about the architecture (including model size), hardware, training compute, dataset construction, training method, or similar. I'm curious whether they have continued to scale up model size/compute significantly or if they have managed to make significant innovati…

I'm assuming they scaled up the model significantly, given the limited availability of the trained model and the increased pricing. Seems like they don't have enough clusters of A100s to go around at the moment.

Re: GPT-4

#518

Post by holtkam2 »

Can someone point me to where I can find information on API pricing for GPT-4? Or, have they not released pricing info yet?

Re: GPT-4

#519

Post by gretch »

manbash wrote: post #324
Imnimo wrote: post #218

A class of problem that GPT-4 appears to still really struggle with is variants of common puzzles. For example: >Suppose I have a cabbage, a goat and a lion, and I need to get them across a river. I have a boat that can only carry myself and a single other item. I am not allowed to leave the cabbage and lion alone together, and I am not allowed to leave the lion and goat alone together. How can I safely get all three…

Honest question: why would you bother expecting it to solve puzzles? It's not a use case for GPT.

The worry here is that GPT has no problem being confidently wrong. A better answer would have been "I can't solve logic problems".

Instead one day, non-technical people will try to use it for all sorts of use cases like legal advice, or medical advice, or advanced math, and it will simply mislead them rather than saying nothing.

Re: GPT-4

#520

Post by jarenmf »

jsheard wrote: post #468
Imnimo wrote: post #218

A class of problem that GPT-4 appears to still really struggle with is variants of common puzzles. For example: >Suppose I have a cabbage, a goat and a lion, and I need to get them across a river. I have a boat that can only carry myself and a single other item. I am not allowed to leave the cabbage and lion alone together, and I am not allowed to leave the lion and goat alone together. How can I safely get all three…

A funny variation on this kind of over-fitting to common trick questions - if you ask it which weighs more, a pound of bricks or a pound of feathers, it will correctly explain that they actually weigh the same amount, one pound. But if you ask it which weighs more, two pounds of bricks or a pound of feathers, the question is similar enough to the trick question that it falls into the same thought process and contorts…

Just tested and GPT4 now solves this correctly, GPT3.5 had a lot of problems with this puzzle even after you explain it several time. One other thing that seem to have improved is that GPT4 is aware of word order. Previously, GPT3.5 could never tell the order of the word in a sentence correctly.
Post reply on HN