Federal 'superintelligence for science' push

The federal government is explicitly shifting funding toward applying large, high‑capability AI systems to accelerate scientific discovery rather than to build frontier models, and is releasing public datasets (190 datasets cited) to train those scientific models. That policy reorients public science dollars toward data and compute‑driven work, with consequences for university funding, research bottlenecks, dual‑use risk, and who sets scientific agendas. — If governments prioritize 'superintelligence for science' they will reshape research incentives, data governance, and national capacity in ways that matter for innovation, security, and equity.