The short-term goal of a tool like this is to sell products. The more ambitious long-term goal is to shift cultural norms, blurring the lines between advertising and reality until the question you're asking is no longer consciously asked. At least, not by average people, and not at the point of purchase.

I find it easy to envision a world, maybe 50 years from now, in which the very concept of "truth in advertising" is viewed as a lost, idyllic fantasy. Something people are nostalgic for, but feel powerless to regain.

As long as the Product Manger gets a bonus/promotion then it's a success.

I've noticed something similar in Facebook marketplace ads for used furniture. Most of the images are AI generated to look like a Pottery Barn catalog, then the last image will be the actual item, full of scratches and other damage, sitting in a messy garage.

Is it gonna be less distorted than just seeing the shirt on a model photographed by a professional in the perfect light?

Sounds exactly like something the marketing department would use to up the convertion rate

I bet if you had a way to easily train a LORA on you trying on various clothing types the models would do pretty well but I agree without that I can't think of a way to get it to work. Reminds me of the flattering mirrors scam https://finance.yahoo.com/news/company-swears-controversial-...

One joy of online shopping, especially for people doing it in an impulsive and/or addictive way (it's not really rare) is the satisfaction they get from the imagination of having it.

An acquaintance of mine was buying many and not wearing most, as she did not attend that many social occasions. Still, she kept buying.

Eventually, she had to face the actual problem in her life that bothered her. She ended up dealing with it, terribly.

If it doesn’t fit, then you must have gained weight while the item was in transit

There's definitely a predatory "everything will look good", but you can also leverage exactly that part for your own purposes - I've done pre-shopping a couple of times by asking for something like "a grid of 9 versions of my photo, wearing different types of X that look good". Definitely helped with choosing a good style.

The more honest ones are “how it looks on you” and don’t promise fit that depends on so many measurements that aren’t even visible in pics.

Sure, it’s idealized, but some people benefit from seeing color / neckline / etc on themselves as a visual reference.

Me, I’m a text-learner so I don’t get it at all. But I know people who get value.

Definitely. I think the appeal to some isn't GenAI per se but image-to-image generation like "make this item look like a talented photographer shot it".

But I can't imagine that your conversion wouldn't benefit by showing actual imperfect photos of your products, especially when you run a platform with 1k new listings per month. I'm personally turned off by something that looks like 3D, especially when all the different colour variations look the exact same.

I use nanobanana 2 to test changes in paint and flooring/tiling to great success. And the real pro move is taking those images to a designer to tweak the remaining 20% or so.

The cynic in me says this is just an advancement and logical continuation from the known problem of sycophantic behavior in text-to-text chatbot format LLM, to image generation models.

To be fair in many cases the actual clothes aren't much better. I've had two pair of the same pants, same brand, same size fit noticeably differently.

Good or bad, I’m pretty sure showing things in the best light has always been the point of “marketing”. There is a reason ads aren’t filled with ugly people with misshaped bodies, and it isn’t because the intention is to reflect reality, so what you allude to as problematic is what a lot of businesses call a feature.

I see the ai product ads and it's pretty funny. The product is shown but is "broken" as far as scale is concerned.

Sort of like how they tried to make gandalf big and the hobbits small in lord of the rings (which wasn't very convincing to me)

User-targeted fashion advice is the polite goal. AI designed to generate images of people is actually racing to capture the entertainment markets. They want to be ready to replace models/actors in everything from fashion mags to porn studios. That is where the money is.

I have a similar use case at work for previewing construction material and such in our catalogue applied to user uploaded images.

Results are mixed, expensive, but it really feels you're few months off the next improvement to really nail it. It's already good enough.

Wonder what Qwen image will provide over nano banana.

If the online retailer has a no quibble returns policy - then at there is at least a strong incentive to minimise returns, rather than oversell.

I've been struggling with this question myself. But isn't this a model training/use problem (a.k.a skill issue )? Isn't there a way to make these models be faithful to how people will actually look?

i'd like it to notice things i would miss, like "this is ring-spun shirt, so it will sit like this on your torso" or "these pleats will require you to iron them" etc

It’s clear to us here on HN but to the average person today the way these models actually work is beyond the realm of constraints and reasoning.

Maybe this will lead to more business for tailors doing alterations, assuming the clothes people end up buying are expensive enough to justify it.

This will push us even more to go outside and visit actual shops. Becouse thanks to AI we will have more time? Will we?

A prompt went viral recently where people were sharing their pictures and asking chatgpt to visualise what their looksmatched partner would look like

The result was always someone extremely good looking

There’s going to be an entirely new class of mental disorders that will emerge from people being deluded by AI

Yet it pushes the goal of the brand/shop forward: make the product more appealing and sell

But they dont have to do either of these things. Thats just what people prompt

> But these models will always make the clothes fit your body and show you in flattering light and so on

This is something that can be fixed over time. And if this forces clothing manufacturers to stick more to their advertised "specs" (width/length), then it's a win for us.

How does this fry pan look with a fish in it? Ask Mr Bean.

[dead]

That's hilarious. These keywords are applied globally, even on pages like https://qwen.ai/usagepolicy, where they very kindly ask you not to use their products for sexual content.

Run this in console to see all the tags:

  document.querySelector('meta[name="keywords"]').content

I assume they use a SEO tool that automatically adds these meta keywords to optimize for some search engines.

It apparently adds common search terms that contain words like "qwen". This evidently includes possibly mistyped searches for "gwen" or "ben" in a NSFW context.

Maybe someone knows more about how such SEO tools work, and where they pull the data from.

Bonus points for also covering misspellings like "pregbant".

What value does the 'keywords' meta tag even have these days? 77+ KB of crap added to the page weight. Web development is full of idiots.

I think the Qwen team is probably aware that the NSFW community is very quick to adopt any new image gen model (see: Civitai). So, it seems like a good SEO approach to try and surface their model in search results that said community is likely already checking.

That is hilarious as porn is illegal in China. But I guess pron SEO is allright, if it is against the west.

Wow you really weren't kidding. Fully automated AI slop tentacles for everyone!

https://pastes.io/uenL6X9K

It also seems to have an obsession with this celebrity, based on how many times ctrl-f for "stefani" turns up a result.

https://en.wikipedia.org/wiki/Gwen_Stefani

Interesting too that they would include "ai friend" when China just added restrictions on AI "partners" such that many services stopped offering them rather than try to adhere to the restrictions.

Brainrot is the new opium war

[dead]

GPT Image 1 ended up with a yellow tint without training on another image-generation model's output. It's just that humans like pictures with a soft sunset glow, and this is a very easy global signal for a preference model to pick up on, and for a image-generation model to imitate. So optimizing for aesthetic appeal makes everything slightly tinted by default, unless you make sure to countersteer.

Maybe it was trained in a Mexican data center https://en.wikipedia.org/wiki/Mexican_filter

yellow/red tint is an extremely common problem not matter the photograph source you train on

source: work at a photograph start, even training on raw images things get tinted, it is an uphill battle

AI-generated images are part of the web now, if you're doing ordinary web scraping, you can't avoid training on generated images.

You clearly haven't met a Chinese RedNote user.

[dead]

that's a very interesting point to test new models. I speak an Indian language called "Tamil" and have always tested new models with Tamil but also have tried little bit of Arabic (quranic verses) with previous GPT image 2 and Nanobanana pro and they have nailed it. Don't know if it was because of extensive training data.

Considering Qwen-Image-2.0 weights have not been released either, it unfortunately looks unlikely.

After Z-Image and the original Qwen-Image - I think they've pivoted completely to closed source. From the images I've seen, Qwen-Image 3.0 is just a subpar equivalent to other proprietary models like gpt-image-2 and nb-pro.

Why release it just so Cursor/Azure/Amazon can profit off of it? Unless OpenAI actually opens it up fat chance.

Krea-2-Turbo. I've even got it working locally on my M5 iPad Pro.

It depends on what you need, but Krea/Klein9b/Ideogram4/Z-Image are among the best right now for text2image and Qwen Edit and Klein are probably still the best at editing.

Not sure about OCR specifically, but the newer (past quarter) vision language models all have a lot of post training on detecting garbled text specifically. You can feed some of the old stable diffusion outputs into a modern model and they can figure out pretty reliably if the text gen is mangled. I think the feature is probably used as part of the RL for the image gen to correct for bad text rendering.

…but the illustrative diagrams are a simulacrum; if you ask Qwen, or any image-generator, for an “accurate” poster-design featuring a representation of a model of an atom and explaining its constituent parts I expect you’ll get an imitation-airbrush rendering of red, blue, and grey table-tennis balls orbiting in perfect circles; you might get an electron-shell diagram if you’re lucky. What you won’t get is anything remotely related to probability-clouds.

Edit: just to test myself I asked Nano Banana 2 to generate “an undergraduate infographic poster about how atoms work” - and the result was something right out of a middle-school science textbook and very Bohr…

My wife has taken all her recipes and fed them through ChatGPT image gen to make zine pages and they’re really cool! She’s building a recipe book for the kids so they’ll know all the recipes from their childhood.

[dead]

Short answer yes.

Slightly longer answer for older text to image models you teach them how to encode images and text into the same latent space. Then you simply do a conversion, take a text input, put it into latent space and then extract the image that latent space represents.

That's exactly how they do it.

There are ML models that do the reverse and output image to text, which assist quite a lot.

The better the text represents the unique thing in the photo, the better the model understands what that text means.

I found this 3b1b guest video on diffusion helpful: https://www.youtube.com/watch?v=iv-5mZ_9CPY

Their site is still broken, months later. Pinching to zoom in progressively covers the content on the page.

Kinda wild to think they haven't fixed it yet. On mobile you cannot zoom in to view their images.

Maybe the site is vibe coded and nobody at Qwen actually knows how to fix it..

Another astonishing UX fail was (still is) on Xiaomi's page[1] - where they try, and successfully fail, to showcase a video editor implemented by their model.

They embed a screen capture hosted on a Chinese hosting site BiliBili, too bad the video is 360p pixel soup.

To be able to change the resolution you are required to:

- know to use desktop view in the browser to make the resolution button show

- know Chinese to understand which button is "resolution" and that you need to login to switch resolution

- have and account with the video hoster or be willing to register

I was curious, but not that curious...

[1] https://mimo.xiaomi.com/mimo-v2-5-pro#AFull-FeaturedVideoEdi...

I can give one innocent example - I made an app that summarises D&D sessions, including images to depict some of the scenes.

But I understand the jaded perspective. I guess the idea of ‘deepfakes’ didn’t give AI image gen a great reputational start.

The right models and LORAs can get rid of it right now. People apparently really like everyone to look like over exposed over filtered mannequins, so the big companies target that look.

There is an IRL phenomena, glass skin skincare and makeup, where the person's skin is evenly flat and toned and glossy. Did you think they are more plasticky than IRL models?

I find it mildly interesting how your comment repeats itself but with different phrasing

"the premium of in person" is a string of words that only means something to people in a very specific circle where everyone is constantly trying to tastemake and write in lower case.

Ironically they have extremely limited influence on how the larger world moves: normal people won't meaningfully change their habits around eating good food or meeting up with each other just because AI can generate a picture of either.

I would attribute any in-person cultural shift to the generation of 20-somethings trapped online during their teenage years from COVID. They are finally old enough, and have enough power to run things.

the era of wabi-sabi is among us.

The moat is the training data. But somehow we've collectively decided that it's not.

Yes, it generates the text without using a font. Same is true of other image models like ChatGPT Images and Gemini Nano Banana and Midjourney.

It's a proprietary model as a service - it doesn't require anything outside of a browser. But even when they released open-weight (the original Qwen-Image) it's a diffusion model and can't use custom font files.

This model isn't supposed to contain all the numerical data. It will give you a (usually) matching graph transformed from one you provide or from a table of information you provide. Or you can pipeline from an LLM doing research on that data first. But expecting an image gen model to get you GDP info has got to be one of the worst possible approaches.

> Especially text rendering

That's true though. I still got some completely fried letters in headings.

Probably because you have too few upvoted comments. The rules likely got stricter when LLMs started to infiltrate Hacker News.