Musk says AI will surpass total human intelligence within 5 years; 40% of global jobs at risk

Tesla CEO Elon Musk predicted that AI will surpass the combined intelligence of all humanity within five years.

In a recently published interview with The Economist, Musk said he believes AI will exceed the sum of human intelligence "in about five years," adding that there will be "nothing AI can't do better than humans, except for being human."

Musk said the most likely outcome is "an age of remarkable abundance where anyone can have whatever they can think of," acknowledging that the idea "may sound absurd."

He added that "of course, something like an all-out global nuclear war could derail that, but assuming no World War III, we are heading toward an age of abundance."

He also acknowledged that the risk of AI and robots veering toward catastrophe is "not zero," but said he has chosen to press ahead with development regardless.

"I don't see a way to actually stop the enormous momentum of AI and robots," Musk said. "Even if there were a stop button, you probably shouldn't press it."

Musk, who leads Tesla in developing the general-purpose humanoid robot Optimus, projected that between 100 million and 1 billion humanoid robots will replace physical labor within five years.

The IMF has estimated that 40 percent of jobs worldwide are exposed to the spread of AI, with that share rising to 60 percent in advanced economies.

Goldman Sachs has estimated that about 300 million full-time jobs globally are at risk of automation.

The World Economic Forum projected that 170 million new jobs will be created by 2030 while 92 million are displaced, for a net gain of 78 million positions.

Geoffrey Hinton, the professor widely known as the "godfather of AI," has forecast mass unemployment. Speaking on the podcast "Diary of a CEO" last year, Hinton said AI "will just replace everyone" when it comes to "routine intellectual labor."

He also challenged the conventional wisdom that AI will create new jobs, saying that "to have a job that AI can't do, you'll need to be very highly skilled."

Mustafa Suleyman, CEO of Microsoft AI, told the Financial Times in May that AI will achieve "human-level performance across most, if not all, professional tasks," and predicted that most work done at a computer will be fully automated within one to 18 months.

Warnings have also emerged from the economics community.

Stanford University's Digital Economy Lab released a statement on July 13 titled "We Must Act Now: A Statement on the Economic Transition from AI," signed by more than 200 people including 16 Nobel laureates in economics. Signatories included Paul Krugman, Niall Ferguson, Tyler Cowen, Reid Hoffman, Eric Schmidt, Jason Furman and Gita Gopinath. The lab warned that AI "could bring about an unprecedented economic transformation larger than the Industrial Revolution, unfolding over a much shorter period," and said it "may bring both risks such as large-scale job displacement and opportunities such as significant improvements in living standards."