Tempus AI CEO says AI development should be slowed down to ensure safety and capital efficiency

Tempus AI CEO Eric Lefkofsky said in a media interview on Monday: "Given that current models are capable of self-training and self-improvement, I can easily imagine a moment when one can't help but wonder: 'Will this get out of control, beyond our grasp?'" He expressed support for slowing down AI development, citing both safety considerations and capital efficiency. He praised the efforts made by Anthropic, OpenAI, and SpaceX in the healthcare sector, and downplayed concerns about "if America’s largest companies slow down investment, the US may fall behind." "I don't think reducing the speed from 100 miles per hour to 70 miles per hour poses any real risk," he said. "I don't think other countries will surpass us because of this."