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

GPT-5

openai.com

Post reply on HN
Print view View original

991–1000 of 1001 posts

Re: GPT-5

#991

Post by andreygrehov »

radu_floricica wrote: post #851
highfrequency wrote: post #331

It is frequently suggested that once one of the AI companies reaches an AGI threshold, they will take off ahead of the rest. It's interesting to note that at least so far, the trend has been the opposite: as time goes on and the models get better, the performance of the different company's gets clustered closer together. Right now GPT-5, Claude Opus, Grok 4, Gemini 2.5 Pro all seem quite good across the board (ie the…

The clustering you see is because they're all optimized for the same benchmarks. In the real world OpenAI is already ahead of the rest, and Grok doesn't even belong in the same group (not that it's not a remarkable achievement to start from scratch and have a working production model in 1-2 years, and integrate it with twitter in a way that works). And Google is Google - kinda hard for them not to be in the top, for…

In my experience, Grok is miles ahead of ChatGPT. I canceled my OpenAI subscription in favor of Grok. I was one of the first OpenAI subscribers.

Re: GPT-5

#992

Post by junon »

junon wrote: post #896

Anecdotal review: Been using it all morning. Had to switch back to 4. 5 has all of the problems that 2/3 had with ignoring any context, flagrantly ignoring the 'spirit' of my requests, and talking to me like I'm a little baby. Not to mention almost all of my prompts result in a several minute wait with "thinking longer about the answer".

Update shortly after my post:

They've removed access to GPT-4 and below. Therefore I've removed their access to my card.

Re: GPT-5

#995

Post by unsupp0rted »

jablongo wrote: post #628
highfrequency wrote: post #331

It is frequently suggested that once one of the AI companies reaches an AGI threshold, they will take off ahead of the rest. It's interesting to note that at least so far, the trend has been the opposite: as time goes on and the models get better, the performance of the different company's gets clustered closer together. Right now GPT-5, Claude Opus, Grok 4, Gemini 2.5 Pro all seem quite good across the board (ie the…

It's also worth considering that past some threshold, it may be very difficult for us as users to discern which model is better. I don't think thats what's going on here, but we should be ready for it. For example, if you are an ELO 1000 chess player would you yourself be able to tell if Magnus Carlson or another grandmaster were better by playing them individually? To the extent that our AGI/SI metrics are based on…

You don't have to be even good at chess to be able to tell when a game is won or lost, most of the time.

I don't need to understand how the AI made the app I asked for or cured my cancer, but it'll be pretty obvious when the app seems to work and the cancer seems to be gone.

I mean, I want to understand how, but I don't need to understand how, in order to benefit from it. Obviously understanding the details would help me evaluate the quality of the solution, but that's an afterthought.

Re: GPT-5

#996

Post by neehao »

highfrequency wrote: post #331

It is frequently suggested that once one of the AI companies reaches an AGI threshold, they will take off ahead of the rest. It's interesting to note that at least so far, the trend has been the opposite: as time goes on and the models get better, the performance of the different company's gets clustered closer together. Right now GPT-5, Claude Opus, Grok 4, Gemini 2.5 Pro all seem quite good across the board (ie the…

three points: 1. i have often wondered about whether rapid tech. progress makes underinvestment more likely.

2. ben evans frequently makes fun of the business value. pretty clear a lot of the models are commodotized.

3. strategically, the winners are platforms where the data are. if you have data in azure, that's where you will use your models. exclusive licensing could pull people to your cloud from on prem. so some gains may go to those companies ...

Re: GPT-5

#997

Post by TheoGone »

highfrequency wrote: post #331

It is frequently suggested that once one of the AI companies reaches an AGI threshold, they will take off ahead of the rest. It's interesting to note that at least so far, the trend has been the opposite: as time goes on and the models get better, the performance of the different company's gets clustered closer together. Right now GPT-5, Claude Opus, Grok 4, Gemini 2.5 Pro all seem quite good across the board (ie the…

Part of it is they all copy and over train, often against the TOS, on each other's models.

Re: GPT-5

#998

Post by TheoGone »

highfrequency wrote: post #331

It is frequently suggested that once one of the AI companies reaches an AGI threshold, they will take off ahead of the rest. It's interesting to note that at least so far, the trend has been the opposite: as time goes on and the models get better, the performance of the different company's gets clustered closer together. Right now GPT-5, Claude Opus, Grok 4, Gemini 2.5 Pro all seem quite good across the board (ie the…

Part of it is the top LLM companies (OpenAI, Mistral) all copy and over train, often against e.g. Claude's or DeepSeek's TOS, on each other's models.

Re: GPT-5

#999

Post by jaybrendansmith »

It does make me wonder about human training, aka education. Could it be that ingesting/reading a better quality literature, math, and other topics would necessarily result in a higher intelligence?

Re: GPT-5

#1000

Post by kenny239 »

highfrequency wrote: post #331

It is frequently suggested that once one of the AI companies reaches an AGI threshold, they will take off ahead of the rest. It's interesting to note that at least so far, the trend has been the opposite: as time goes on and the models get better, the performance of the different company's gets clustered closer together. Right now GPT-5, Claude Opus, Grok 4, Gemini 2.5 Pro all seem quite good across the board (ie the…

not a researcher for long enough....but we are witnessing open source effort & Chinese models starting to fall one "level" behind the most advanced models, mainly due to a lack of compute i think...

on the other hand, there are still some flaws regarding GPT-5. for example, when i use it for research it often needs multiple prompts to get the topic i truly want and sometimes it can feed me false information. so the reasoning part is not fully there yet?

Post reply on HN