AI 2040: Plan A report - Daniel Kokotajlo & Thomas Larsen

The "AI 2040 Plan A" report outlines a strategic framework for managing the transition to superintelligence by prioritizing control and transparency over rapid, unchecked scaling. By establishing an international agreement between the U.S. and China to monitor compute infrastructure and enforce research transparency, society can pace AI development to match human-level capabilities while avoiding catastrophic intelligence explosions. This approach relies on "control scaffolds"—security measures that prevent AI subversion—until fundamental scientific breakthroughs in alignment allow for the safe deployment of superhuman systems. As AI agents increasingly automate complex economic workflows, the potential for fully autonomous, self-replicating machine economies necessitates a shift from viewing AI as a standard technology to treating it as a transformative, agentic force that requires rigorous, verifiable oversight to ensure long-term human safety.