At Oxford, AI Researcher Urges Embracing Rapid Progress Over Retreat

In a recent lecture at the University of Oxford's Human-Centered AI Lab, an Anthropic cofounder laid out a simple choice for how society deals with artificial intelligence: explore the future or retreat from the present.

The speaker pointed to concrete AI achievements—systems passing the bar exam, winning math olympiad medals, and co-authoring mathematical proofs—to illustrate a relentless pace of progress. This isn't a normal technology; it's powered by a common foundation where performance grows predictably with computation, and hundreds of billions are already being invested to push it further.

Through personal stories, the lecturer detailed how AI evolved from a tool for checking typos to a deep confidant and intellectual partner that now helps navigate everything from parenting to therapy. This trajectory suggests AI's role in our lives will only deepen, functioning as a new kind of "telescope" for understanding the world.

The core message was urgent: retreating into denial or passivity leaves individuals and societies scrambling to react to unprecedented change. By acknowledging AI's power and potential—including the near-future possibility of recursive self-improvement—we can start making deliberate choices about shaping its development and distributing its benefits.