Interpersonal AI
Humans & AI
Interpersonal AI
Affective Use of AI
youtube.com2.9% of people that use Claude.ai use it for conversations on affective topics
‘AI alignment’ - how the model should behave
Amanda Askell - Anthropic
“How a morally motivated kind human would act”
Ethics is a lot more like physics than we think - uncertainty, hypotheses
Models should be uncertain about moral systems
“Single inference forward path way” - what does that mean?
Interacting to ‘lean in’ rather ‘leaning out’ of connection with people
What makes a good life?
In 1938, Harvard set out to answer that question by launching the longest scientific study on human happiness ever conducted. Eight decades, three generations, and thousands of data points later, the results are in and they might just change how you live. #psychology #motivation #motivationalquotes
instagram.comCan LLMs do the same for people?
Ideas related to this collection