In 25 years we'll reminisce on the times when we could find a human artist who wouldn't impose Google's or OpenAI's rules on their output.

the open-source models will catch up, 100%

Open models don't seem to be catching up the LLM-based image gen at this point.

ChatGPT's imagegen has been released for half a year but there isn't anything remotely similar to it in the open weight realm.

Give it another 50 years. Or maybe 10. Or 5? But there's no way it won't catch up.

Are you asking it to recreate people?

No, and no nudity, no reference images. Example: 'athlete wearing a health tracker under a fitted training top'

Can you give us an example?

'athlete wearing a health tracker under a fitted training top'

Failed to generate content: permission denied. Please try again.

It's not the censorship safeguard. Permission denied means you need a paid API key to use it. It's confusing, I know.

If you triggered the safeguard it'll give you the typical "sorry, I can't..." LLM response.

Have some examples?