Import AI 456: RSI and Economic Growth; Radical Optionality for AI Regulation; Neural Computer

Researchers from the Institute for Law & AI propose 'radical optionality' for AI regulation, arguing governments should invest in tools and institutions now to preserve decision-making ability in future crises. They recommend transparency and reporting requirements, whistleblower protections, information-sharing within and between governments, flexible rules, assessments, improved security of model weights, and increased hiring of technical talent. A paper from Meta and KAIST introduces the Neural Computer (NC), a neural system that unifies computation, memory, and I/O in a learned runtime state, with early prototypes using CLI and GUI interfaces. Economists from Forethought, Columbia University, and the University of Virginia model that recursive self-improvement (RSI) via AI automation could cause explosive economic growth, finding that 13% automation across all sectors or full automation of software R&D with 5% elsewhere could tip the economy into a singularity within about six years, with hardware research being the dominant lever.