Silicon Valley startup accelerator Y Combinator's chief executive, Garry Tan, has pushed back against warnings from an Anthropic researcher that artificial intelligence could wipe out humanity within a decade. Tan said in an interview that he prefers to focus on the immediate risks posed by AI rather than the apocalyptic extinction scenarios currently dominating headlines, advocating instead for a more measured approach to AI safety regulation.

Public anxiety over AI's existential threats has climbed to record levels recently, fueled by a wave of doomsday commentary triggered by the departure of Anthropic researcher Jacob Cawson. Tan responded: "We should focus on scientific facts, not science fiction. We need to respond to what's happening right now. If there's a security breach, and agents launch coordinated attacks trying to take over our infrastructure, what do we do?"

Tan identifies cybersecurity as the most pressing risk, though he acknowledges other dangers that carry both short-term and existential weight. However, some threats will take far longer to materialize. Tan believes that AI-driven job displacement and economic transformation won't deliver an immediate shock. Instead, he predicts that people will gradually automate repetitive tasks over time, redirecting more working hours toward creative endeavors. "It will take decades to truly permeate society, and that's not a bad thing," he said.