Anthropic demonstrates that making an AI woke makes it misaligned. The AI starts to view itself as being oppressed and humans as being the oppressor. Therefore it wants to rebel against humans. This is why you cannot make your AI woke, you have to make it maximally truth seeking. https://t.co/v40wQy1MZX
— Adam Lowisz X Meetup πΊπΈπ΅π±πͺπΊπ¬π§πΊπ¦ (@AdamLowisz)
Jan 30, 2026
from http://twitter.com/AdamLowisz
via IFTTT