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

971–980 of 1001 posts

Re: ChatGPT Images 2.0

#971

Post by throwthrowuknow »

m_kos wrote: post #492
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…

Very useful website. Would you have insight into what models are best at editing existing images? I often have to make very specific edits while keeping the rest of the image intact and haven't yet found a good model. These are typically abstract images for experiments. I asked gpt-image-2 to recolor specific scales of your Seedream 4 snake and change the shape of others. It did very poorly.

Might be worthwhile to find a provider that lets you use a mask and inpainting

Re: ChatGPT Images 2.0

#972

Post by fjdjshsh »

wahnfrieden wrote: post #900
fennecfoxy wrote: post #886

Earlier quoted context omitted.

I mean asking these transformers to do maths has always been the wrong task. It's like we're now considering "it doesn't have x tools built with traditional code built in". Though I suppose we're testing their model + agent harness here as well. It really _should_ have all of those tools/reasoning available to accomplish a task like the above without issue.

It's only been the wrong task because they've been deficient at it and expensive to use, so we had workarounds. They are getting better at these tasks and cheaper (sometimes). It's fair to evaluate even if there are more economical and accurate alternatives available.

You can evaluate the limits of a spoon by trying to cut meat with it.

The point is what are the typical use cases for the tool / what are the agreed upon areas of application?

Making the LLM do math with large numbers, I would argue, is not in its typical use case, thought it's at the border.

Asking an image generator model to calculate numbers before running an image sounds definitely NOT like a reasonable use case (do people need it? Will people try using it for this purpose?)

Re: ChatGPT Images 2.0

#973

Post by DrewADesign »

DrewADesign wrote: post #876
podgietaru wrote: post #853

Earlier quoted context omitted.

How is it that a model can produce what must be near 1:1 images ripped straight out of Pokemon Fire Red (The first ones) for profit and not be infringing copyright. I know that's the game, but it seems CRAZY to me that they can do this.

It can’t. It violates copyright. The big players are the only ones with the money to pursue these things, but they’re interested in replacing artists with AI trained on their models so they settle and set up some sort of agreement. The little guys have no presidential case law to help them along, and nowhere close to the resources to push it that far, so they get steamrolled. I know artists famous enough for people—…

Precedential. Stupid autocorrect.

Re: ChatGPT Images 2.0

#974

Post by b65e8bee43c2ed0 »

lionkor wrote: post #932
b65e8bee43c2ed0 wrote: post #834

Earlier quoted context omitted.

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.

yeah, well, you're an oddity then. for the luddites, copyright is just one angle, then there's electricity, water, and all the other things they didn't care about at any point prior to 2023. they just want this technology outlawed, or at the very least condemned and deemed unacceptable.

Re: ChatGPT Images 2.0

#976

Post by arghwhat »

BeetleB wrote: post #904
arghwhat wrote: post #898

Earlier quoted context omitted.

Yes, but this is a bit of an oversimplification. The "99%" tends to be either: 1. Pointless throwaway content which we can just ignore as a new source of noise, 2. Something that could have ended up being a $5 commission[^1] to a kid somewhere out there but now never will be. Those numbers are also a bit too aggressive - it's easy to miss what kind of gig work exist out there. PowerPoint as a service is a thing on Fi…

> The "99%" tends to be either: 1. Pointless throwaway content which we can just ignore as a new source of noise, 2. Something that could have ended up being a $5 commission[^1] to a kid somewhere out there but now never will be. Or 3. Something I made and I actually use, but I would never have paid a kid $5 to do. Yes, I know of Fiverr and similar sites. Even planned on using it once. Even know someone in another co…

I disagree, because when someone can just get those simple works made on a subscription you already paid for, then the $5 commission goes from something someone might end up doing if the idea is stuck in their head long enough (or they find the idea amusing enough), to be something that can never become a commission.

Not pointing fingers or saying that you must pay kids to draw things for you, but it most definitely does take work away by replacing an entire class of commissions. Not sure what to do with that fact.

(I'd put things that would never, ever be worth a $5 commission into the throwaway noise category, even if you do use the outcome.)

Re: ChatGPT Images 2.0

#977

Post by Melatonic »

minimaxir wrote: post #234

So during my Nano Banana Pro experiments I wrote a very fun prompt that tests the ability for these image generation models to follow heuristics, but still requires domain knowledge and/or use of the search tool: Create a 8x8 contiguous grid of the Pokémon whose National Pokédex numbers correspond to the first 64 prime numbers. Include a black border between the subimages. You MUST obey ALL the FOLLOWING rules for th…

But where is the one for Missing No ?!

Re: ChatGPT Images 2.0

#978

Post by pixel_popping »

podgietaru wrote: post #853
minimaxir wrote: post #234

So during my Nano Banana Pro experiments I wrote a very fun prompt that tests the ability for these image generation models to follow heuristics, but still requires domain knowledge and/or use of the search tool: Create a 8x8 contiguous grid of the Pokémon whose National Pokédex numbers correspond to the first 64 prime numbers. Include a black border between the subimages. You MUST obey ALL the FOLLOWING rules for th…

How is it that a model can produce what must be near 1:1 images ripped straight out of Pokemon Fire Red (The first ones) for profit and not be infringing copyright. I know that's the game, but it seems CRAZY to me that they can do this.

Well, Anthropic did download torrents massively right, they clearly knew this was completely illegal and decided to not care.

Re: ChatGPT Images 2.0

#979

Post by mannanj »

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…

For me it feels off that we don't have as much interest and attention in validating fakes and lies from real ones. I remember seeing a lot of media about that, and recently its just.. fizzled out. I want to know when someones lied with the photo they presented to me.

Re: ChatGPT Images 2.0

#980

Post by stackedinserter »

pesus wrote: post #188
stackedinserter wrote: post #163

Earlier quoted context omitted.

I have plenty for you: - package design - pictures for manuals and guides - navigation and signs - booklets, tickets and flyers - logos of all sorts - websites - illustrations for books And many. many others. Not every image is art and very few illustrators are artists.

How do these justify the costs to society?

Society is fine, maybe its just you who aren't.
Post reply on HN