this post was submitted on 25 Feb 2024
150 points (77.4% liked)

Technology

59312 readers
5006 users here now

This is a most excellent place for technology news and articles.


Our Rules


  1. Follow the lemmy.world rules.
  2. Only tech related content.
  3. Be excellent to each another!
  4. Mod approved content bots can post up to 10 articles per day.
  5. Threads asking for personal tech support may be deleted.
  6. Politics threads may be removed.
  7. No memes allowed as posts, OK to post as comments.
  8. Only approved bots from the list below, to ask if your bot can be added please contact us.
  9. Check for duplicates before posting, duplicates may be removed

Approved Bots


founded 1 year ago
MODERATORS
you are viewing a single comment's thread
view the rest of the comments
[–] [email protected] 16 points 8 months ago (2 children)

The very same AI that shows pictures of black people with dreadlocks when asked to show a typical viking also has a braindead response for a question involving Hitler and a guy who posts retarded tweets and regularly pisses off his shareholders? I am shocked.

AI is still so ridiculously tainted by bias and the relative infancy of the tech.

[–] [email protected] 28 points 8 months ago* (last edited 8 months ago) (2 children)

Time to shelve retarded as an insult. I used it too in the past, but it's best not to use it any more.

Intellectually handicapped folks can't help it and didn't choose it.

[–] [email protected] 7 points 8 months ago* (last edited 8 months ago) (2 children)

Wait, doesn't it just say "removed" for all of you? Is my instance doing that?

[–] [email protected] 4 points 8 months ago (1 children)
[–] [email protected] 5 points 8 months ago (2 children)

This might actually make me switch instances. I don't get what the point of the filter even is, I'm not 6 anymore, it's fine if I see no-no words.

[–] [email protected] 2 points 8 months ago

I see the original word. I’m on sh.itjust.works if that’s helpful, you can imagine they probably don’t filter bad words :-)

[–] [email protected] 2 points 8 months ago

My biggest complaint is that I can't actually see what slur was removed, and there are some softball words in the list that I honestly don't give a shit about, so I am always hesitant to report one of these posts because I don't know exactly what they said.

"Removed" could be the n-word, or it could just be a rude word, so how am I supposed to make a call whenever I see it?

[–] [email protected] 4 points 8 months ago (1 children)

Yeah... It's the slur filter, now days it can be enabled or not by the instance but in past it was hard coded. Unfortunately I don't think there is individual user setting to disable it for themselves.

[–] [email protected] 3 points 8 months ago (1 children)

What a load of bs. Why make it a blanket decision at all???

[–] [email protected] 1 points 8 months ago (1 children)

They didn't even want to allow people disabling it at all ... like you needed to fork it and change the code if you wanted that ... If anything it has improved.

[–] [email protected] 1 points 8 months ago

At least something >_>

[–] [email protected] 4 points 8 months ago

While agree that Gemini does a terrible job on this task, the model that generates the images (text-to-image) is entirely different from the model that incorrectly answers the comparison between Musk and Hitler (text-to-text). All they share is the same product name (Gemini) and the same company that developed them. Probably entirely different teams built these two models.