hahaha, these people are so cooked
Conversation
Notices
-
Embed this notice
Ein unbehaarter Affe (lechimp@mastodon.pub.solar)'s status on Friday, 06-Feb-2026 06:11:43 JST
Ein unbehaarter Affe
-
Embed this notice
Rich Felker (dalias@hachyderm.io)'s status on Friday, 06-Feb-2026 06:11:42 JST
Rich Felker
@lechimp I'm not clear what you're trying to say. From the abstract, it looks like there are patterns of language usage that are well correlated between descriptions of the behaviors of the people making "AI" and of abusive parents, which leads to these statements being emitted from the slop machines. It doesn't say anything about the machines but it might be insightful about the people making them.
-
Embed this notice
Ein unbehaarter Affe (lechimp@mastodon.pub.solar)'s status on Friday, 06-Feb-2026 12:05:07 JST
Ein unbehaarter Affe
@dalias they got fooled into thinking that there are psychological states of mind behind the words. they made psychological assessments of some pareidolia and published a paper (or at least an abstract to a paper) talking about them as humans.
-
Embed this notice
Ein unbehaarter Affe (lechimp@mastodon.pub.solar)'s status on Friday, 06-Feb-2026 12:05:08 JST
Ein unbehaarter Affe
@dalias i mean, yes, this would be one, fairly generous, interpretation of the abstract. what i rather see is highly anthropomorphizing language, starting in the title “[…] Reveal Internal Conflict in Frontier Models”. so i very much think these people indeed ascribe these “multi-morbid synthetic psychopatholog[ies]” to the models themselves and not to their makers. and i think its bonkers, because they “argue that these responses go beyond role-play”.
Rich Felker repeated this. -
Embed this notice
Rich Felker (dalias@hachyderm.io)'s status on Friday, 06-Feb-2026 13:50:00 JST
Rich Felker
@lechimp Yeah attributing agency to them is utter bullshit. The way they "go beyond roleplay" is by revealing interesting statistical correlations in language usage.
-
Embed this notice