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

791–800 of 1001 posts

Re: GPT-5

#791

Post by alexey-salmin »

EthanHeilman wrote: post #786
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 is frequently suggested that once one of the AI companies reaches an AGI threshold, they will take off ahead of the rest. This seems to be a result of using overly simplistic models of progress. A company makes a breakthrough, the next breakthrough requires exploring many more paths. It is much easier to catch up than find a breakthrough. Even if you get lucky and find the next breakthrough before everyone catch…

Supposedly the idea was, once you get closer to AGI it starts to explore these breakthrough paths for you providing a positive feedback loop. Hence the expected exponential explosion in power.

But yes, so far it feels like we are in the latter stages of the innovation S-curve for transformer-based architectures. The exponent may be out there but it probably requires jumping onto a new S-curve.

Re: GPT-5

#793

Post by ipnon »

Telemakhos wrote: post #734

I am thoroughly unimpressed by GPT-5. It still can't compose iambic trimeters in ancient Greek with a proper penthemimeral cæsura, and it insists on providing totally incorrect scansion of the flawed lines it does compose. I corrected its metrical sins twice, which sent it into "thinking" mode until it finally returned a "Reasoning failed" error. There is no intelligence here: it's still just giving plausible output.…

This is a great test because it’s something you could teach an elementary school kid in an hour.

Re: GPT-5

#794

Post by didibus »

some-guy wrote: post #283
minimaxir wrote: post #41

The marketing copy and the current livestream appear tautological: "it's better because it's better." Not much explanation yet why GPT-5 warrants a major version bump. As usual, the model (and potentially OpenAI as a whole) will depend on output vibe checks.

As someone who tries to push the limits of hard coding tasks (mainly refactoring old codebases) to LLMs with not much improvement since the last round of models, I'm finding that we are hitting the reduction of rate of improvement on the S-curve of quality. Obviously getting the same quality cheaper would be huge, but the quality of the output day to day isn't noticeable to me.

Agree, I think they'll need to move to performance now. If a model was comparable to Claude 4, but took like 500ms or less per edit. A quicker feedback loop would be a big improvement.

Re: GPT-5

#795

Post by techpineapple »

podgietaru wrote: post #562
sundarurfriend wrote: post #317

Some people have hypothesized that GPT-5 is actually about cost reduction and internal optimization for OpenAI, since there doesn't seem to be much of a leap forward, but another element that they seem to have focused on that'll probably make a huge difference to "normal" (non-tech) users is making precise and specifically worded prompts less necessary. They've mentioned improvements in that aspects a few times now,…

I just don’t know that you’d name that 5. The jump from 3 to 4 was huge. There was an expectation for similar outputs here. Making it cheaper is a good goal - certainly - but they needed a huge marketing win too.

Did they really have another choice? if no big leap was on the horizon are they just never going to release 5? I mean, from a marketing perspective.

Re: GPT-5

#796

Post by Davidzheng »

thegeomaster wrote: post #70

SWE-Bench Verified score, with thinking, ties Opus 4.1 without thinking. AIME scores do not appear too impressive at first glance. They are downplaying benchmarks heavily in the live stream. This was the lab that has been flexing benchmarks as headline figures since forever. This is a product-focused update. There is no significant jump in raw intelligence or agentic behavior against SOTA.

what does it mean for a bench to be not impressive when it's saturated?

Re: GPT-5

#797

Post by throwpoaster »

I’ve been working on an electrochemistry project, with several models but mostly o3-pro.

GPT-5 refused to continue the conversation because it was worried about potential weapons applications, so we gave the business to the other models.

Disappointing.

Re: GPT-5

#798

Post by tejohnso »

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…

I think the expectation is that it will be very close until one team reaches beyond the threshold. Then even if that team is only one month ahead, they will always be one month ahead in terms of time to catch up, but in terms of performance at a particular time their lead will continue to extend. So users will use the winner's tools, or use tools that are inferior by many orders of magnitude.

This assumes an infinite potential for improvement though. It's also possible that the winner maxes out after threshold day plus one week, and then everyone hits the same limit within a relatively short time.

Re: GPT-5

#800

Post by xnx »

surround wrote: post #297

GPT-5 knowledge cutoff: Sep 30, 2024 (10 months before release). Compare that to Gemini 2.5 Pro knowledge cutoff: Jan 2025 (3 months before release) Claude Opus 4.1: knowledge cutoff: Mar 2025 (4 months before release) https://platform.openai.com/docs/models/compare https://deepmind.google/models/gemini/pro/ https://docs.anthropic.com/en/docs/about-claude/models/overv...

Does this indicate that OpenAI had a very long pretraining process for GPT5?
Post reply on HN