this post was submitted on 21 May 2025
19 points (77.1% liked)
ChatGPT
9722 readers
3 users here now
Unofficial ChatGPT community to discuss anything ChatGPT
founded 2 years ago
MODERATORS
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
Less than two years ago AI couldn't make an image of a person with the right amount of fingers.
Now we are here. In two years AI video could produce a TV show or movie.
It was fast, for sure, but tv shows in two years are very unlikely. If we could get at least more consistent results and better character control, that would already be a lot
AI has been able to make pictures with the right amount of fingers for about four years. It involved cherry picking and sometimes inpainting. The biggest thing that’s changed is resolution and prompt adherence.
Even the best models sometimes (frequently) give people extra fingers.