Relevant: AIs at the current capability level may be important for future safety work

I'm going to make a bid for keeping digital minds in scope

We need more AI governance explainers with a fancy British accent

The AI Action Summit in France could play an important role in building upon the safety agreements of the Bletchley Park and Seoul Summits, or it could fail to make any meaningful contributions to safety. I'm not personally familiar with the team but it seems to have good experience with French AI governance.