this post was submitted on 06 Aug 2025
689 points (84.1% liked)
Technology
73878 readers
3577 users here now
This is a most excellent place for technology news and articles.
Our Rules
- Follow the lemmy.world rules.
- Only tech related news or articles.
- Be excellent to each other!
- Mod approved content bots can post up to 10 articles per day.
- Threads asking for personal tech support may be deleted.
- Politics threads may be removed.
- No memes allowed as posts, OK to post as comments.
- Only approved bots from the list below, this includes using AI responses and summaries. To ask if your bot can be added please contact a mod.
- Check for duplicates before posting, duplicates may be removed
- Accounts 7 days and younger will have their posts automatically removed.
Approved Bots
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
I was talking about ai training on ai output, ai requires genuine data, having a feedback loop makes models regress, see how ai makes yellow pictures because of the ghibli ai thing
Sure, that mainly applies when it's the same model training on itself. If a model trains on a different one, it might retrieve some good features from it, but the bad sides as well
AI requires genuine data, period. Go read about it instead of spewing nonsense.
If they weren’t trained on the same data, it ends up similar
Training inferior models with superior models output can lower the gap between both. It’ll not be optimal by any means and you might fuck its future learning, but it will work to an extent
The data you feed it should be good quality though