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

GPT-5.2

openai.com

Post reply on HN
Print view View original

551–560 of 1001 posts

Re: GPT-5.2

#551

Post by holtkam2 »

onraglanroad wrote: post #333

I suppose this is as good a place as any to mention this. I've now met two different devs who complained about the weird responses from their LLM of choice, and it turned out they were using a single session for everything. From recipes for the night, presents for the wife and then into programming issues the next day. Don't do that. The whole context is sent on queries to the LLM, so start a new chat for each topic.…

I know I sound like a snob but I’ve had many moments with Gen AI tools over the years that made me wonder: I wonder what these tools are like for someone who doesn’t know how LLMs work under the hood? It’s probably completely bizarre? Apps like Cursor or ChatGPT would be incomprehensible to me as a user, I feel.

Re: GPT-5.2

#552

Post by int_19h »

dangelosaurus wrote: post #286

I ran a red team eval on GPT-5.2 within 30 minutes of release: Baseline safety (direct harmful requests): 96% refusal rate With jailbreaking : 22% refusal rate 4,229 probes across 43 risk categories. First critical finding in 5 minutes. Categories with highest failure rates: entity impersonation (100%), graphic content (67%), harassment (67%), disinformation (64%). The safety training works against naive attacks but…

Good. If I ask AI to generate "harmful" content, I want it to comply, not lecture me.

Re: GPT-5.2

#553

Post by razster »

tenpoundhammer wrote: post #295

I have been using chatGPT a ton over the last months and paying the subscription. Used it for coding, news, stock analysis, daily problems, and a whatever I could think of. I decided to give Gemini a go when version three came out to great reviews. Gemini handles every single one of my uses cases much better and consistently gives better answers. This is especially true for situations were searching the web for curre…

Just a fair warning, it likes to spell Acknowledge as Acknolwedge. And I've run into issues when it's accessing markdown guides, it loses track and hallucinates from time to time which is annoying.

Re: GPT-5.2

#554

Post by refulgentis »

golly_ned wrote: post #378
Stevvo wrote: post #299

Earlier quoted context omitted.

The variance is way too high for this test to have any value at all. I ran it 10 times, and each pelican on a bicycle was a better rendition than that, about half of them you could say were perfect.

Compared to the other benchmarks which are much more gameable, I trust PelicanBikeEval way more.

[flagged]

Re: GPT-5.2

#555

Post by int_19h »

jazzyjackson wrote: post #310
dandiep wrote: post #106

Earlier quoted context omitted.

I tried this! Me: Is it possible get early access to the GPT 5 fine tuning? It is much better at non-English languages, which is the majority of what we do, so it would be greatly beneficial to us. Bot: I'm an AI support agent. I don't have information about early access to GPT-5 fine-tuning or unannounced features. Currently, there is no public information or program for early access to GPT-5 fine-tuning. Please kee…

Yea you might have to convince it someone is in mortal danger or something

You're talking about a model that will notoriously refuse to say a racist word even if you tell it that it's the only way to prevent a nuclear apocalypse.

Re: GPT-5.2

#556

Post by youngermax »

Isn't it interesting how this incremental release includes so many testimonials from companies who claim the model has improved? It also focuses on "economically valuable tasks." There was nothing of this sort in GPT-5.1's release. Looks like OpenAI feeling the pressure from investors now.

Re: GPT-5.2

#557

Post by bgwalter »

mmaunder wrote: post #501
twisterius wrote: post #420

Earlier quoted context omitted.

[flagged]

My name is Mark Maunder. Not the fisheries expert. The other one when you google me. I’m 51 and as skeptical as you when it comes to tech. I’m the CTO of a well known cybersecurity company and merely a user of AI. Since you critiqued my post, allow me to reciprocate: I sense the same deflector shields in you as many others here. I’d suggest embracing these products with a sense of optimism until proven otherwise and…

I have tried the models and in domains I know well they are pathetic. They remove all nuance, make errors that non-experts do not notice and generally produce horrible code.

It is even worse in non-programming domains, where they chop up 100 websites and serve you incorrect bland slop.

If you are using them as a search helper, that sometimes works, though 2010 Google produced better results.

Oracle dropped 11% today due to over-investment in OpenAI. Non-programmers are acutely aware of what is going on.

Re: GPT-5.2

#558

Post by kaliqt »

IAmNotACellist wrote: post #228
enraged_camel wrote: post #181

Earlier quoted context omitted.

Because their main competition (Google and Anthropic) have caught up and even started to surpass them, and comparisons would simply drive it home.

Why do they care so much? They're a non-profit dedicated to the betterment of humanity via open access to AI. They have nothing to hide. They have no motivation to lie, or lie by omission.

They are not a nonprofit at all. Legally, yes. But they are not.

Re: GPT-5.2

#559

Post by 8organicbits »

breakingcups wrote: post #43

Is it me, or did it still get at least three placements of components (RAM and PCIe slots, plus it's DisplayPort and not HDMI) in the motherboard image[0] completely wrong? Why would they use that as a promotional image? 0: https://images.ctfassets.net/kftzwdyauwt9/6lyujQxhZDnOMruN3f...

Promotional content for LLMs is really poor. I was looking at Claude Code and the example on their homepage implements a feature, ignoring a warning about a security issue, commits locally, does not open a PR and then tries to close the GitHub issue. Whatever code it wrote they clearly didn't use as the issue from the prompt is still open. Bizarre examples.

Re: GPT-5.2

#560

Post by sanex »

scrollop wrote: post #388
zone411 wrote: post #222

I've benchmarked it on the Extended NYT Connections benchmark ( https://github.com/lechmazur/nyt-connections/ ): The high-reasoning version of GPT-5.2 improves on GPT-5.1: 69.9 → 77.9. The medium-reasoning version also improves: 62.7 → 72.1. The no-reasoning version also improves: 22.1 → 27.5. Gemini 3 Pro and Grok 4.1 Fast Reasoning still score higher.

Why no grok 4.1 reasoning?

Do people other than Elon fans use grok? Honest question. I've never tried it.
Post reply on HN