Reverse game theory aims to control powerful AI models by designing rules that give them incentives to follow instructions. Standard game theory predicts how rational actors behave under given rules. Reverse game theory works the other way: it designs the rules to produce the behavior people want. The report's example is cake splitting. If one person cuts the cake and the other chooses first, the cutter has an incentive to make the portions equal.
Researchers believe the approach could align AI agents more closely with human rules and instructions. The report says big questions remain about whether these mechanisms will work in practice.
The same newsletter reports that President Trump signed an executive order at the end of September directing government agencies to rebrand artificial intelligence as "Super Intelligence." The report says the new term is slowly appearing on government websites and documents, but has not caught on across the board. It says some large tech companies still use "AI." The report also raises a concern that the term could become a political loyalty test, with people who use it seen as signaling allegiance to the administration and those who don't becoming targets for attacks.
The newsletter also lists other stories, including the US midterm elections as a test of AI's influence on politics, Nvidia telling data center clients to limit business with China, and hackers turning supermarket fridges into a