A CERN for AI-assisted science (terrytao.wordpress.com)

A call to action has emerged from the recent developments in AI-assisted science, particularly following OpenAI's solution to the Navier–Stokes Millennium Prize problem. Dimitris Koukoulopoulos argues for the establishment of a publicly funded "CERN for AI-assisted science" to address the growing concerns over access inequality in AI tools for research. Highlighting the significant risks presented by the control of advanced AI systems by a few private companies, he emphasizes the need for a collective academic response that fosters equitable access to powerful AI technologies essential for scientific inquiry. This initiative would provide researchers with user-friendly interfaces to robust AI systems while ensuring strong confidentiality protocols and accountability for sensitive data. Koukoulopoulos envisions a collaborative framework, supported by public investment from entities like the EU and Canada, where scientists could access advanced AI resources without the restrictive barriers posed by commercial interests. By prioritizing public AI infrastructure, the scientific community can safeguard its independence, nurture early-career researchers, and promote a more inclusive environment for innovation and research, ensuring that the future of science remains driven by collective knowledge rather than corporate agendas.