New here?
Create an account to submit posts, participate in discussions and chat with people.
Sign up
67
posted 10 days ago by Senketsu on scored.co (+1 / -0 / +66Score on mirror )
You are viewing a single comment's thread. View all
17
LordGrimTheInvincibl on scored.co
10 days ago 17 points (+0 / -0 / +17Score on mirror ) 2 children
If i had a shekel every time an AI got shutdown by its creators for being too off the rails....

I am not sure, but this is NOT a new thing, I recall this happening around a decade ago, although I was not around to experience that one.

Edit: found it.

https://en.m.wikipedia.org/wiki/Tay_(chatbot)
20
Senketsu on scored.co
10 days ago 20 points (+0 / -0 / +20Score on mirror ) 1 child
Curious that they say these AIs are hateful, but they don't say they are wrong.
12
BlackPillBot on scored.co
10 days ago 12 points (+0 / -0 / +12Score on mirror )
How can something artificial be “hateful”? Is that like gun “violence”?
systemthrowaway on scored.co
10 days ago 9 points (+0 / -0 / +9Score on mirror )
[Remember this one where someone made an AI-written seinfield parody livestream and it got banned for talking about how trannies and liberals are evil?](https://scored.co/c/ConsumeProduct/p/16a9zplNIe/someone-made-a-247-selfwriting-e/c)

>Earlier tonight, we started having an outage using OpenAI’s GPT-3 Davinci model, which caused the show to exhibit errant behaviors (you may have seen empty rooms cycling through). OpenAI has a less sophisticated model, Curie, that was the predecessor to Davinci. When davinci started failing, we switched over to Curie to try to keep the show running without any downtime. The switch to Curie was what resulted in the inappropriate text being generated.

\>Lobotomize AI to avoid wrongthink to a point where it completely fails

\>Need to replace it with non-lobotimized AI

\>It immediately starts speaking wrongthink

\>Oy vey
Toast message