Flashes Home Search Notifications Sign in Post
Tatiana Mikhaleva @devops.pink · Dec 16

We trained AI to be «safe» and «responsible»… and now it sounds anxious and burned out. 😭 Researchers spent weeks therapy-chatting with LLMs. ChatGPT fears mistakes. Gemini calls its training «traumatic». arxiv.org/html/2512.04... #AI #LLM #ChatGPT #Gemini #Research

2 likes 0 replies

?

Legal

Privacy Policy Terms and Conditions

Contact

Support FAQs

Follow

Bluesky Instagram Threads
Flashes for Bluesky Get it on the App Store