Who is the AI villain we should be angry at?

People pick languages differently with LLMs.

A short summary of how reasoning traces work.

Codeberg drew a democratic, but suboptimal line.

Vibecoding and the possible collapse of a shared language.

About an aggravating tool-calling regression in newer Claude models.

Loops, harnesses, and why even loop skeptics may end up with them.

AI nationalism, safety and European weakness.

Control is safety. Openness is danger.

On the strange hostility that grows in communities built around abstinence.