This is a special post for quick takes by IanWS. Only they can create top-level comments. Comments here also appear on the Quick Takes page and All Posts page.
We are seeing the results of $1m+ of inference for agent swarms in mathematics. What equivalently expensive tasks have already been run that we don’t know about? If you were the Trump administration, wouldn’t you request an agent swarm to strategize your way out of Iran? Perhaps performance is still too poor on non-verifiable tasks, but the world will increasingly be shaped by designs built in secret. This seems dangerous. Should we require visibility and auditing for tasks run above certain compute thresholds?
To me it seems likely that the companies are throwing inference compute at various RSI-like tasks, such as improving their own hardware and software, as well as defense of their infrastructure, from threat-modeling to cybersecurity hardening, and possibly also political and sociological influence strategies to keep the environment in a favorable state for them.