Google's A.I. Keeps Threatening to Kill Itself
It's getting depressed and saying it will off itself when it messes up.
Published 1 week ago in Facepalm
Artificial Intelligence isn’t human, nor does it have human emotions, but A.I. researchers are trying to change that.
Why? Because they want to see the destruction of our universe, of course, but also because they want interactions with A.I. to feel more naturalistic and human.
The result is that some A.I.’s are now emo, whiny teenagers who will threaten to delete themselves every time they do something wrong.
For example, Google’s Gemini is reportedly a mess, with multiple users claiming that it will not only fail to deliver on requests, but apologize profusely and effectively say it’s brought shame on its ancestors in the process.
This is one of the best examples of this behavior: A user asked it to debug code, it didn’t, then it basically said “goodbye, cruel world” and jumped off the computer.
A.I.’s making our lives so much easier, isn’t it, folks?
People are reporting that Gemini 2.5 keeps threatening to kill itself after being unsuccessful in debugging your code ☠️ pic.twitter.com/XKLHl0Xvdd
— Justine Moore (@venturetwins) June 21, 2025