Elon Musk's Economist Interview: AI Surpassing Humans in 5 Years, OpenAI's Unintended Race, and a Peer-Review Safety Plan

AI Will Surpass Human Intelligence — Within Five Years

Musk made his most specific timeline prediction yet: AI systems may exceed the sum of all human intelligence within roughly five years . He elaborated that, in that timeframe, "there really won't be anything that AI can't do better than humans, apart from being human, perhaps" . Looking further ahead, he projected that within ten years, humans will no longer be in control, and robots will usher in an era of "amazing abundance" where the concept of money becomes meaningless .

A Peer-Review Safety Solution for Rival AI Models

Musk proposed that competing AI companies should be allowed to review each other's models for safety risks before they are launched . He described this as a peer-review system for frontier AI — analogous to scientific peer review in academia — where labs examine one another's leading models to identify and mitigate dangers before public release . The idea was presented as an industry-wide safety collaboration mechanism, though The Economist and other outlets noted skepticism about its feasibility .

The 10–20% Extinction Risk and 'Enjoying the Ride'

Despite his dire timelines, Musk did not walk back his long-stated estimate that there is a 10–20% probability that AI could cause human extinction . Instead, he said he had reached a "philosophical conclusion" to focus on the positive, urging people to essentially "enjoy the ride" because the momentum behind AI and robotics is now unstoppable .

Key caveat: As The Economist itself noted, Musk's predictions should be taken with a measure of skepticism given his uneven track record . The BBC's reporting captured his tone as "let's enjoy the ride" — a remarkable posture for someone who simultaneously warns of a 10–20% chance of human annihilation .