Fengjiao

where are you building them? How much does it cost to build?

Thank youuu we’ll definitely do!

You can use it for free and I can give you unlimited usage, although not licensing for now! Didn't see your email. Send me an email at [email protected]

I'm a woman and ex-creator on an app like this. AI matters as a productivity tool as all content platforms struggle with getting more content, but yes there's a bunch of other important things

Thanks for the insightful comment. All the above is true! C.C is an artist we recruited from a Chinese manhua app. We're reproducing and repackaging some killer tropes in Asian markets for the west

contact info & identifiers, same as most apps (tiktok, instagram etc)

We use LoRA models with stable diffusion models. They can be trained quickly with minimal 30-50 images of your subject. (https://civitai.com/articles/2099/lora-models-and-how-to-use...)

we actually removed most of the tracking because they're not necessary, should be updated in a new release

Thank you for checking out the app - more will be there! We thought we'd launch early and learn what people like before committing to too many projects

It’s webcomic romance stories. And the-billionaire-spends-all-his-money-on-me romance is a popular trope, out of others. There’s long been a market for this for late night binge read. 50 shades of gray also came out of this market. It is a bit different from the Pulitzer and NYT bestsellers.

You might also underestimate how many of your favorite writers ghostwrite stories like this for a side income.

It’s a tool for human creators. The stories are written by writers we found from platforms like Wattpad, Kindle and Chapters. Then the art is made by artists using the AI Studio where they control the characters’ pose and facial expressions. That’s the production process.

So if this works, all creators get profit. It should enable more human creators to make money, since previously they couldn’t complete a finished product alone.

G point - kinda did that for HN so people don't expect Marvel stuff, but might be overdone

bit buzz wordy yes. The actual speed up is now 1 artist does what used to be 1 main artist + 1 line artist + 1 colorist + postprocessing. So they can make a nice finished product alone.

Great point - thanks!! Yes, we should put better material out there.

The real value would be to help these creators make money from making comics, which is not known as the most money grabbing career, so the shortbread app site is consumer focused and like you said, should attract them to read the content

We struggled quite a bit trying to create an attractive non-white male lead, and ended up with this face that was stable. In most stable diffusion models trained on OSS datasets, keywords like "middle-age African man" "30 year old Asian man" will deform the face and head shape somewhat

Sorry for the notification issues. Not intended like this - will fix now!

It's a preview link that adapts to your phone screen. Our artists use this link to render the content in realtime as they create. Published stuff goes through a microservice and is standardized.

Just getting started so the content is limited

I'm a woman. Female. Female engineer. I have a vagina and I write code.

All our creators, writers and artists are women.

There are two comic episodes in the post. You can open: https://create.shortbread.ai/viewer/3d25370c-a4eb-4df4-af8e-... https://create.shortbread.ai/viewer/866c2873-9962-4ebe-8f0c-...

The ideal creator is someone who understands comics storytelling, but lacks the time or drawing skills. So either a writer or artist can do. In general, we see artists with prior drawing skills use it better because they understand good character acting, emotions and cinematography, etc.

Thank youu - on it!

LMAO

Yes was talking to a creator about this. Getting the eyeliner to match actually uplifts the story a lot. I imagine some simple rotation/flipping image post processing can help a lot here

Hope it helps!

Do you mean the text in the speech bubbles? Those are generated by the AI and we fill them in the bubbles. If you’re seeing text in images like a billboard, it probably isn’t that great quality. I heard Dalle3 does pretty good text in image!

hue hue hue (evil founder laugh)

Ty just fixed it. whew