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

ChatGPT Images 2.0

openai.com

Post reply on HN
Print view View original

931–940 of 1001 posts

Re: ChatGPT Images 2.0

#931

Post by lionkor »

unsupp0rted wrote: post #766
lionkor wrote: post #613

Every cent you spend on this, remember: The people who made this possible are not even getting a millionth of a cent for every billion USD made with it (they are getting nothing). Same with code; that code you spent years pouring over, fixing, etc. is now how these companies make so much money and get so much investment. It's like open source, except you get shafted.

This doesn’t bother me one bit. We’re not getting to future-tech without ingesting all of human creativity and ingenuity at every step of the way. Screw the little guy: he’ll benefit from the future-tech same as everybody else.

Or not, it doesn't really matter and nobody seems to care. This isn't for the good of humanity, there's no indication that it is

Re: ChatGPT Images 2.0

#932

Post by lionkor »

b65e8bee43c2ed0 wrote: post #834
lionkor wrote: post #613

Every cent you spend on this, remember: The people who made this possible are not even getting a millionth of a cent for every billion USD made with it (they are getting nothing). Same with code; that code you spent years pouring over, fixing, etc. is now how these companies make so much money and get so much investment. It's like open source, except you get shafted.

let's not pretend that it would make any difference if these models were trained with only licensed data. you folx would still decry this technology. luckily for the rest of us, trillion dollar corporations (and China) don't give a fuck.

Weird hypothetical. Not sure who "you folx" is.

For me, that would solve my issue with it.

Re: ChatGPT Images 2.0

#933

Post by lionkor »

bradley13 wrote: post #808
lionkor wrote: post #687

Earlier quoted context omitted.

That's a very strange view. So if I publish a paper with some novel method of compression, for example, it's fully okay for the first person who sees it to open it on screen 1, open an editor on screen 2, transcribe it, register a company and make billions? Is that how you WANT the world to work? Because that sure isn't how it works, and that's not been how it works, that's not been legal, and your argument is to sud…

Your comment is way off-base. If you publish a paper, the expression is copyrighted, but your algorithm is not protected at all. If you want to protect the algorithm, you need a patent. Then, the person "making billions" needs to pay you a license fee. However, even then: - An algorithm is not patentable. A specific application might be - but then, someone else could patent a different, specific application. - If you…

You're right, I was wrong with my example.

> The fact that AI is more efficient at this? So what? That does not in any way affect the principle.

Well it's not a human, so exceptions for humans shouldn't apply

Re: ChatGPT Images 2.0

#934

Post by minimaxir »

Palmik wrote: post #882
minimaxir wrote: post #858

Earlier quoted context omitted.

Huh, that is indeed better. If ChatGPT Images 2.0/gpt-2-image is more nondeterministic than usual, than that is in itself a useful data point.

Did you enable thinking for your experiment? Are you sure you were on the 2.0 rather than 1.5 version?

That experiment image was directly through the API on high. (no Thinking parameter like the Web UI)

Re: ChatGPT Images 2.0

#935

Post by vunderba »

valarauko wrote: post #897
vunderba wrote: post #83

OpenAI’s gpt-image-1.5 and Google’s NB2 have been pretty much neck and neck on my comparison site which focuses heavily on prompt adherence, with both hovering around a 70% success rate on the prompts for generative and editing capabilities. With the caveat being that Gemini has always had the edge in terms of visual fidelity. That being said, gpt-image-1.5 was a big leap in visual quality for OpenAI and eliminated m…

That's lovely. My own personal benchmark has been to ask the various models to generate a functional pair of novelty New Year's Eve glasses on a person, that don't just plonk the year onto the top of regular frames.

Thanks. That's a good one~ Lens type stuff that involves reflections/refraction is a neat challenge for generative models. I did some editing tests that involved replacing an apartment window with a mirror back when Nano-Banana Pro was released and was rather stunned by the results.

https://mordenstar.com/blog/edits-with-nanobanana/#through-t...

Re: ChatGPT Images 2.0

#936

Post by zulban »

Fraterkes wrote: post #627
zulban wrote: post #499

Earlier quoted context omitted.

Maybe reread my comment. Would you not want to see a mount Everest sized Lego cat? Even if it were my cat? Again - your quip sounds good but when you think about it, it's flatly wrong.

This doesn't make sense, if I want to see a lego-cat slopimage I can just prompt a model myself (and have it be of my own cat). There's no reason for you to be involved in any part of that process, because the point of this stuff is that you are not doing anything.

The claim is that people don't / shouldn't want to see something if humans can't be bothered to make it. I provided a counter example. So the claim is nonsense.

Re: ChatGPT Images 2.0

#937

Post by cubefox »

hahahacorn wrote: post #60

One of the images in the blog ( https://images.ctfassets.net/kftzwdyauwt9/4d5dizAOajLfAXkGZ7... ) is a carbon copy of an image from an article posted Mar 27, 2026 with credits given to an individual: https://www.cornellsun.com/article/2026/03/cornell-accepts-5... Was this an oversight? Or did their new image generation model generate an image that was essentially a copy of an existing image?

I checked all the images on the blog post and I'm quite sure that the one you talk about isn't there.

Re: ChatGPT Images 2.0

#938

Post by gusmally »

vunderba wrote: post #83

OpenAI’s gpt-image-1.5 and Google’s NB2 have been pretty much neck and neck on my comparison site which focuses heavily on prompt adherence, with both hovering around a 70% success rate on the prompts for generative and editing capabilities. With the caveat being that Gemini has always had the edge in terms of visual fidelity. That being said, gpt-image-1.5 was a big leap in visual quality for OpenAI and eliminated m…

Such a fun site, thank you! I was surprised that Seedream4 passed the mermaid test since it's hard to tell whether they are in the water or submerged, and the mermaid has something funny going on with her left hand.

Re: ChatGPT Images 2.0

#939

Post by vunderba »

gusmally wrote: post #938
vunderba wrote: post #83

OpenAI’s gpt-image-1.5 and Google’s NB2 have been pretty much neck and neck on my comparison site which focuses heavily on prompt adherence, with both hovering around a 70% success rate on the prompts for generative and editing capabilities. With the caveat being that Gemini has always had the edge in terms of visual fidelity. That being said, gpt-image-1.5 was a big leap in visual quality for OpenAI and eliminated m…

Such a fun site, thank you! I was surprised that Seedream4 passed the mermaid test since it's hard to tell whether they are in the water or submerged, and the mermaid has something funny going on with her left hand.

Yeah seedream's attempt does have a bit of an uncanny valley effect: the mermaid/dolphin are only partially submerged, but there’s water above them with sunlight reflecting on the surface, and the mermaid’s hand looks disconnected from the angle of her arm.

That’s why I gave it a bronze. To me, it falls into that “barely passing” category, similar to Gemini 2.5 Flash Image on that test. Seedream also took a major hit to its weighted score because of how many attempts it took to get something even remotely passable out of it.

Thanks for the feedback!

Re: ChatGPT Images 2.0

#940

Post by linen »

parasti wrote: post #554

A great technical achievement, for sure, but this is kind of the moment where it enters uncanny valley to me. The promo reel on the website makes it feel like humans doing incredible things (background music intentionally evokes that emotion), but it's a slideshow of computer generatated images attempting to replicate the amazing things that humans do. It's just crazy to look at those images and have to consciously r…

Why are so many on HN unable to see through the B.S. and hype? Everything in the trailer feels unvaried and derivative. It does text and filters well (grit/grain, UI etc) but all the posters, comics, and infographics feel the same. They've all got matching structure and color palettes and once you've seen enough of them, you can easily spot them in a crowd. I'm not sure why people are falling for this, the AI voices in the trailer are ridiculous too.
Post reply on HN