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

Tell HN: Dont use Claude Design, lost access to my projects after unsubscribing

news.ycombinator.com

Post reply on HN
Print view View original

91–100 of 102 posts

Re: Tell HN: Dont use Claude Design, lost access to my projects after unsubscribing

#95

Post by simonw »

embedding-shape wrote: post #86
semiquaver wrote: post #13

Earlier quoted context omitted.

What do you mean LLMs are blind? All frontier models are multimodal, which means they literally consume images as tokens. They can “see” exactly as well as they can “read”. Also, GPT-Image-2 is not a diffusion model, it is based on Transformers, like other LLMs are.

> Also, GPT-Image-2 is not a diffusion model, it is based on Transformers, like other LLMs are. Where are you getting this from btw? AFAIK, OpenAI hasn't openly talked about what exactly is powering the Images 2.0 stuff, unless I missed something? I think they've said it's not a diffusion model, but I'm not sure they've said what they're doing instead, have they?

I believe it's an evolution of the technique used in GPT-Image-1 (or whatever they called that), which was derived from their work on making GPT-4o an "omni" model that can directly output images and audio in addition to text.

The 2024 GPT-4o launch post https://openai.com/index/hello-gpt-4o/ hints about how that works:

"With GPT‑4o, we trained a single new model end-to-end across text, vision, and audio, meaning that all inputs and outputs are processed by the same neural network."

Re: Tell HN: Dont use Claude Design, lost access to my projects after unsubscribing

#96

Post by embedding-shape »

simonw wrote: post #95
embedding-shape wrote: post #86

Earlier quoted context omitted.

> Also, GPT-Image-2 is not a diffusion model, it is based on Transformers, like other LLMs are. Where are you getting this from btw? AFAIK, OpenAI hasn't openly talked about what exactly is powering the Images 2.0 stuff, unless I missed something? I think they've said it's not a diffusion model, but I'm not sure they've said what they're doing instead, have they?

I believe it's an evolution of the technique used in GPT-Image-1 (or whatever they called that), which was derived from their work on making GPT-4o an "omni" model that can directly output images and audio in addition to text. The 2024 GPT-4o launch post https://openai.com/index/hello-gpt-4o/ hints about how that works: "With GPT‑4o, we trained a single new model end-to-end across text, vision, and audio, meaning tha…

Yeah, that's my belief as well, but haven't seen any concrete explanations about how it works, just the marketing/press releases sadly.

Re: Tell HN: Dont use Claude Design, lost access to my projects after unsubscribing

#97

Post by inquirerGeneral »

embedding-shape wrote: post #48
semiquaver wrote: post #13

Earlier quoted context omitted.

What do you mean LLMs are blind? All frontier models are multimodal, which means they literally consume images as tokens. They can “see” exactly as well as they can “read”. Also, GPT-Image-2 is not a diffusion model, it is based on Transformers, like other LLMs are.

I guess they do "see" but more like "see an explanation of the image", not "see" as in experience visually. They're really bad at details and perfection when it comes to images, and doesn't understand things like visual hierarchy, affordances and other fundamental design concepts. Most of them are able to describe those things with letters, but doesn't seem to actually fundamentally grasp it when asking it to do UIs…

[flagged]

Re: Tell HN: Dont use Claude Design, lost access to my projects after unsubscribing

#98

Post by Topfi »

trq_ wrote: post #64

Hi there, Thariq from the Claude team here. Sorry this is happening, we'll fix it ASAP. We don't want anyone to feel locked into the tool. Claude's designs are HTML/CSS/JS that any editor can handle; we'll make sure it's possible to download them even after you unsubscribe.

Don't know why you were initially downed (maybe some thought you were an impersonator), so vouched for your comment.

Happy to see you on here and great to hear that you'll address this. As mentioned in my comment, data export let's users get access as it stands, but any improvement in UX is always welcome. Maybe making Design accessible via API credits for occasional use might be something you could bring up too.

Thanks for your efforts as an in between the user base and Anthropic. Based on prior personal experience and purely looking in from the outside, I wouldn't be surprising if it can be very challenging and stressful to be in such a position, especially when one may not directly state or say what they think is best in a given situation, so thank you for dealing with a not always very pleasant position where the situation and information you are provided with can change without your say, yet you may be the one to that will be considered responsible in the eyes of the public for a decision you neither made nor could prevent.

Re: Tell HN: Dont use Claude Design, lost access to my projects after unsubscribing

#100

Post by satvikpendem »

TurdF3rguson wrote: post #63
satvikpendem wrote: post #26

Earlier quoted context omitted.

Or just use Google's Stitch, it integrates both code via Gemini and image UI generation via Nano Banana which I'd argue is even better than OpenAI's image models.

It's really not, gpt-image-2 is #1 by over 100 ELO.

What's the source of that, are there image benchmarks?
Post reply on HN