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

GPT-4.5

openai.com

Post reply on HN
Print view View original

161–170 of 1001 posts

Re: GPT-4.5

#161

Post by uh_uh »

Bjorkbat wrote: post #42
Topfi wrote: post #24

Considering both this blog post and the livestream demos, I am underwhelmed. Having just finished the stream, I had a real "was that all" moment, which on one hand shows how spoiled I've gotten by new models impressing me, but on another feels like OpenAI really struggles to stay ahead of their competitors. What has been shown feels like it could be achieved using a custom system prompt on older versions of OpenAIs m…

My first thought seeing this and looking at benchmarks was that if it wasn’t for reasoning, then either pundits would be saying we’ve hit a plateau, or at the very least OpenAI is clearly in 2nd place to Anthropic in model performance. Of course we don’t live in such a world, but I thought of this nonetheless because for all the connotations that come with a 4.5 moniker this is kind of underwhelming.

Pundits were saying that deep learning has hit a plateau even before the LLM boom.

Re: GPT-4.5

#162

Post by pawelduda »

bhouston wrote: post #7

A bit better at coding than ChatGPT 4o but not better than o3-mini - there is a chart near the bottom of the page that is easy to overlook: - ChatGPT 4.5 on AWS Bench verified: 38.0% - ChatGPT 4o on AWS Bench verified: 30.7% - OpenAI o3-mini on AWS Bench verified: 61.0% BTW Anthropic Claude 3.7 is better than o3-mini at coding at around 62-70% [1]. This means that I'll stick with Claude 3.7 for the time being for my…

Does the benchmark reflect your opinion on 3.7? I've been using 3.7 via Cursor and it's noticeably worse than 3.5. I've heard using the standalone model works fine, didn't get a chance to try it yet though.

Re: GPT-4.5

#163

Post by eightysixfour »

sebastiennight wrote: post #154
eightysixfour wrote: post #141

Earlier quoted context omitted.

From Sam's twitter: > After that, a top goal for us is to unify o-series models and GPT-series models by creating systems that can use all our tools, know when to think for a long time or not, and generally be useful for a very wide range of tasks. > In both ChatGPT and our API, we will release GPT-5 as a system that integrates a lot of our technology, including o3. We will no longer ship o3 as a standalone model. Yo…

Ah, great point. Yes, the wording here would imply that they're basically planning on building scaffolding around multiple models instead of having one more capable Swiss Army Knife model. I would feel a bit bummed if GPT-5 turned out not to be a model, but rather a "product".

For me it depends on how the models are glued together. Connected by function calling and APIs? Probably meh...

Somehow working together in the same latent space? That could be neat.

Re: GPT-4.5

#164

Post by lysace »

mjburgess wrote: post #44
camwhite wrote: post #3

I can't wait for fireship.io and the comment section here to tell me what to think about this

You appear to have the direction of causation reversed. (In that fireship does the same)

I wonder if fireship reaction video scripts to AI models based on HN comments can be automated using said AI models.

Re: GPT-4.5

#165

Post by emseetech »

jstummbillig wrote: post #127
serjester wrote: post #58

Earlier quoted context omitted.

I suppose this was their final hurrah after two failed attempts at training GPT-5 with the traditional pre-training paradigm. Just confirms reasoning models are the only way forward.

What it confirms, I think, is, that we are going to need a lot more chips.

Or, possibly, we're stuck waiting for another theoretical breakthrough before real progress is made.

Re: GPT-4.5

#166

Post by rakejake »

ls_stats wrote: post #153
wewewedxfgdf wrote: post #106

I feel like OpenAI is pursuing AGI when Anthropic/Claude is pursuing making AI awesome for practical things like coding. I only ever using OpenAI's coding now as a double check against Claude. Does OpenAI have their eyes on the ball?

>I feel like OpenAI is pursuing AGI I don't think so, the "AGI guy" was Ilya Sutskever, he is gone, he wanted to make OpenAI "less comercial", AGI is just a buzzword for Altmann.

Right. A good chunk of the "old guard" is now gone - Ilya to SSI, Mira and a bunch of others to a new venture called Thinking Machines, Alec Radford etc. Remains to be seen if OpenAI will be the leader or if other players catch up.

Re: GPT-4.5

#167

Post by jampa »

First impression of GPT-4.5:

1. It is very very slow, for some applications where you want real time interactions is just not viable, the text attached below took 7s to generate with 4o, but 46s with GPT4.5

2. The style it writes is way better: it keeps the tone you ask and makes better improvements on the flow. One of my biggest complaints with 4o is that you want for your content to be more casual and accessible but GPT / DeepSeek wants to write like Shakespeare did.

Some comparisons on a book draft: GPT4o (left) and GPT4.5 (green). I also adjusted the spacing around the paragraphs, to better diff match. I still am wary of using ChatGPT to help me write, even with GPT 4.5, but the improvement is very noticeable.

https://i.imgur.com/ogalyE0.png

Re: GPT-4.5

#168

Post by riwsky »

swatcoder wrote: post #57
zaptrem wrote: post #26

GPT 4.5 pricing is insane: Price Input: $75.00 / 1M tokens Cached input: $37.50 / 1M tokens Output: $150.00 / 1M tokens GPT 4o pricing for comparison: Price Input: $2.50 / 1M tokens Cached input: $1.25 / 1M tokens Output: $10.00 / 1M tokens It sounds like it's so expensive and the difference in usefulness is so lacking(?) they're not even gonna keep serving it in the API for long: > GPT‑4.5 is a very large and comput…

> We look forward to learning more about its strengths, capabilities, and potential applications in real-world settings. If GPT‑4.5 delivers unique value for your use case, your feedback (opens in a new window) will play an important role in guiding our decision. "We don't really know what this is good for, but spent a lot of money and time making it and are under intense pressure to announce new things right now. If…

Said the quiet part out loud! Or as we say these days, “transparently exposed the chain of thought tokens”.

Re: GPT-4.5

#169

Post by mchusma »

wow, openai really missed here. Reading the blog I thought like a minor, incremental minor catch up release for 4o. I thought "wow maybe this is cheaper than 4o so it will offset the pricing difference between this and something like Claude Sonnet 3.7 or Gemini 2.0 Flash both of which performs better. But its like 20x-100x more expensive!

In other words, these performance stats with Gemini 2.0 Flash pricing looks reasonable. At these prices, zero usecases for anyone I think. This is a dead on arrival model.

Re: GPT-4.5

#170

Post by taytus »

synapsomorphy wrote: post #155

Claude 3.6 (new 3.5) and 3.7 non-reasoning are much better at pretty much everything, and much cheaper. What's Anthropic's secret sauce?

They ship more focused on their mission than OpenAI.
Post reply on HN