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

ChatGPT Images 2.5

openai.com

Post reply on HN
Print view View original

261–270 of 474 posts

Re: ChatGPT Images 2.5

#261

Post by embedding-shape »

bko wrote: post #217
ks2048 wrote: post #208

Earlier quoted context omitted.

If I eat at a restaurant and that restaurant has images of their food, I want those to be “real images” of what they will bring me when I pay. Otherwise it’s misleading.

I don't think I've ever eaten at a restaurant with real images of their food, pre or post AI.

Not sure exactly what images we're talking about here, places above "medium-scale" typically don't have images on their menus at all, and mostly have real iamges on their Instagram and website, of course professionally taken and edited though.

Re: ChatGPT Images 2.5

#263

Post by strangecasts »

WarmWash wrote: post #245
magicalist wrote: post #216

Earlier quoted context omitted.

> Then will we continue to see the slop accusations? If it doesn't look like the food you're about to purchase, yes? Why is this even a question in this context.

"Here is a picture of the dish, make it look nicer and use that as a reference"

"Users will voluntarily prompt better to avoid misleading people" will be the "people can simply write memory safe C" of the late 2020s

Re: ChatGPT Images 2.5

#264

Post by cbovis »

villish wrote: post #184
bluegatty wrote: post #144

Earlier quoted context omitted.

This an oddly cynical comment, that I think says more about the observer than the observed. People are having fun making images. Like they do all sorts of things on the web. Making an image is as simple as visiting a web-page. That's it.

I hate it because I hate having to look at these AI generated images. In fact I want an internet where I have to explicitly opt-in to see AI generated content in general.

I would imagine a lot of it is mundane shit like getting ChatGPT to render the same room with different coloured walls or provide inspiration on plant pots. I have a decent amount of image usage and it's all things like this, not generating content I intend to distribute.

Re: ChatGPT Images 2.5

#265

Post by dinfinity »

well_ackshually wrote: post #249
bluegatty wrote: post #144

Earlier quoted context omitted.

This an oddly cynical comment, that I think says more about the observer than the observed. People are having fun making images. Like they do all sorts of things on the web. Making an image is as simple as visiting a web-page. That's it.

>People are having fun making images. 3 billion images per week isn't people, it's automated farms, dogshit marketing companies and other attacks on your brain.

All of it?

What percentage is benign?

Do you have a source?

Re: ChatGPT Images 2.5

#266

Post by SirMaster »

Starlevel004 wrote: post #118
meerita wrote: post #111

I've really enjoyed using AI to generate images. For example, Long time ago, I read an amazing five-book saga called Riverworld, and I used AI to recreate many scenes, places, and ideas from the story. Seeing the books come to life through hundreds of images was a great experience. Definitely one of the coolest things to enjoy in 2026. Try it with your favorite books.

[flagged]

I feel like the reason I dislike reading or find reading boring is because I can't. But it's just a theory. Meanwhile I absolutely love movies.

Re: ChatGPT Images 2.5

#267

Post by dinfinity »

rustystump wrote: post #214
bluegatty wrote: post #144

Earlier quoted context omitted.

This an oddly cynical comment, that I think says more about the observer than the observed. People are having fun making images. Like they do all sorts of things on the web. Making an image is as simple as visiting a web-page. That's it.

this is an oddly dismissive comment that says more about the commenter then what is commented on. generating an ai image is vastly more costly than loading a webpage…but still not that costly compared to most other things. this is depressing in the same way content slop or fyp feeds are fun. if you want a silly pic with ur kids, just pay someone on fiver. it will be only slightly more expensive and likely far more ex…

> if you want a silly pic with ur kids, just pay someone on fiver.

Incredible alternative. Talk about depressing.

Re: ChatGPT Images 2.5

#268

Post by nxc18 »

tyjen wrote: post #36
coffeebeqn wrote: post #16

Earlier quoted context omitted.

I see a lot of the slop posters in real life now and I hate them

I think every restaurant on Uber Eats in my area uses AI slop or, minimally, AI "enhanced" food pictures.

I see a ton of AI generated images and descriptions on DoorDash. I get the sense the restaurant is not doing it themselves, I’m not sure if they have any control over disabling it or perhaps they are opting in.

Re: ChatGPT Images 2.5

#269

Post by embedding-shape »

cobolcomesback wrote: post #59
AyanamiKaine wrote: post #12

Omg, I love how the first examples just show how easy you can fake things. Fake being at a party with your friends. Didn't make your bed, no problem, just fake it. The sad part is my mother would love "remixing" my old child photos of me.

Seriously, are these really the best examples they could come up with? What is the point of having a fake picture of your dog in a costume? What’s the point of having a fake picture about being at a party? The only use case I can think of for this is for someone who likes to make up stories and lie about what they’ve done. Is that really the target market?

> The only use case I can think of for this is for someone who likes to make up stories and lie about what they’ve done. Is that really the target market?

As far as I can tell, the biggest uses of these sort of models is to pump out lots of content on multimedia-based social media, so yeah, it's basically for the sort of person who doesn't shy away from exaggerating, omitting or outright lying in order to get more views.

Re: ChatGPT Images 2.5

#270

Post by raincole »

jjcm wrote: post #253
raincole wrote: post #243

Earlier quoted context omitted.

I'm confused. Isn't diffui using its own model?

It's both. I have several models loaded into diffui. Which one each prompt uses is determined based on user preference over time. Whichever image is currently at the top of any image node stack marks a win for the model that generated it, I assign each model an ELO score based on that, and I bias the chance each model is selected based on their ELO score. Right now gpt-image-2 is better than my own model, and it serv…

Oh, you're the dev of diffui? It's completely off-topic, but I've notice that RMB -> Download image will download an .png but it's actually a .webp and it can cause issues (e.g. file explorer doesn't show thumbnail correctly).
Post reply on HN