| ▲ | guzik 11 hours ago |
| Cool, but it's still unusable for me. Somehow all my prompts are violating the rules, huh? |
|
| ▲ | gdulli 11 hours ago | parent | next [-] |
| 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. |
| |
| ▲ | guzik 11 hours ago | parent [-] | | the open-source models will catch up, 100% | | |
| ▲ | raincole 11 hours ago | parent [-] | | 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. | | |
| ▲ | recursive 8 hours ago | parent [-] | | Give it another 50 years. Or maybe 10. Or 5? But there's no way it won't catch up. |
|
|
|
|
| ▲ | mudkipdev 11 hours ago | parent | prev | next [-] |
| Are you asking it to recreate people? |
| |
| ▲ | guzik 11 hours ago | parent [-] | | No, and no nudity, no reference images. Example: 'athlete wearing a health tracker under a fitted training top' |
|
|
| ▲ | Filligree 11 hours ago | parent | prev | next [-] |
| Can you give us an example? |
| |
| ▲ | guzik 11 hours ago | parent [-] | | 'athlete wearing a health tracker under a fitted training top' Failed to generate content: permission denied. Please try again. | | |
| ▲ | raincole 11 hours ago | parent [-] | | 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. |
|
|
|
| ▲ | ASinclair 11 hours ago | parent | prev [-] |
| Have some examples? |