Photo by Aliaksandr Trafimovich/Getty Images
Knowing what challenges advanced AI may pose is necessary but not sufficient to confront them. Researchers in the Center for the Geopolitics of AGI are asking what needs to be planned for now: what practical safeguards, if any, could constrain a superintelligent system? What “no regret” investments could societies benefit from today, regardless of whether a new risk materializes? Across these questions, a common finding emerges: the tools and plans that would be needed in a crisis must be developed long before that crisis arrives.
