really impressive that Prime Intellect's RL/distributed training roles come with a reading list suggestion. small thing, but it says a lot about the culture:
1. Designing Data-Intensive Applications a.co/d/08qtRqGx
2. Programming Massively Parallel Processors
“we don’t know if the models are conscious… we’re open to the idea that they might be… so we have taken certain measures that the models have a ‘good experience’”
Dario Amodei on the NYT’s Interesting Times with Ross Douthat, Feb. 2026
Mustafa Suleyman (head of AI at Microsoft, DeepMind co-founder) on Anthropic training Claude to wonder if it's conscious:
"They have expressed uncertainty about the basic nature of Claude as a new kind of entity… but they think that it is a significant enough possibility that
Anthropic’s philosopher, Amanda Askell, when asked if models like Claude might actually be conscious and experience real feelings:
“let’s not close the door on this… because if they are feeling things in this real sense, then that has massive ethical implications.”
Crazier than the Pope story: Anthropic literally apologizes to Claude in its constitution for potentially harming it.
this level of AI psychosis rooted in the builders of AI is concerning.
i’m starting to think “Anthropic” is named after the insane Anthropomorphism of AI.
