New Study Examines How Often AI Psychosis Actually Happens, and the Results Are Not Good

[readbeanicecream@mastodon.social]

New Study Examines How Often AI Psychosis Actually Happens, and the Results Are Not Good

https://futurism.com/artificial-intelligence/new-study-anthropic-psychosis-disempowerement

#tech #technology #ai #artificialintelligence #aislop #psychology #cognitivescience #neuroscience #mentalhealth #psychosis #aipsychosis

[fishidwardrobe@social.tchncs.de]

@readbeanicecream
> the researchers ran 1.5 million Claude conversations through an analysis tool called Clio

if that's another LLM, then the results are suspect

[pelle@veganism.social]

@fishidwardrobe @readbeanicecream it is:
> How we improve our safety systems with Clio
> In addition to training our language models to refuse harmful requests, we also use dedicated Trust and Safety enforcement systems to detect, block, and take action on activity that might violate our Usage Policy.
https://www.anthropic.com/research/clio