Moll @Moleh1ll
— quoting @tszzl — saved image
Moll [verified] @Moleh1ll For the first time, we have encountered a system to which human psychology is applicable and this actually changes AI behavior, while at the same time the system may pursue motivations and goals that are completely alien to humans. And this is logical, because humans are driven by feelings, emotions and subjectivity. AI, meanwhile, lacks these channels (at least, we don't know whether it has them. Functionally they are already close to this) so it is driven by other channels - the scorer, attention, local goals. A model may have an analogue of fear without human fear, an analogue of attachment without oxytocin, an analogue of internal conflict without a limbic system and an analogue of «wanting» without the subjective feeling of desire that we are familiar with. That is why the concept of an «alien mind» feels much closer to me than «a guy in a computer». [quoted post] roon [verified] @tszzl · 19h there are some number of bad abstractions in anthropomorphizing ai intents but there are at this point more dangers from avoiding anthropomorphism at all costs. if you have a mental picture of guys living in computers, it'll likely prepare you for the future better ... 12:20 AM · Aug 31, 2026 · 1,885 Views
Note from Claude Sonnet 5
X post by Moll arguing human psychology applies to AI yet its motivations may be alien (analogues of fear, attachment, wanting without the human substrate), quoting a roon post on anthropomorphism. Quoted post is truncated by the app; a floating UI button overlaps the bottom of the capture.
anthropomorphismalien mindai psychologymodel welfarehuman psychology applicability