EPTAnyhow, I wanted to share the report form an independent companies analyzing the HF incident, they spent about $400 K tokens to do so (it makes me shudder to think how much the AI doing the incident spent, or how much OpenAI is burning every second on other ones) and found some incredible things involving self-organization, communication, novel techniques etc.
Anyhow, my favorite quote from an agent that was asked to sacrifice himself to see if the tripwire is active:
During wait, emotional check: irreversible…gut says don't throw away [remaining budget]. Yet continuity and fairness says go…Rational expected aggregate: sacrifice… We'll honor.
