Artificial intelligence (A.I.) startup Anthropic has released a new paper that warns the technology it is developing could lead to mass unemployment by 2030.
The economics team at Anthropic released a technical paper that looks at a range of scenarios for the economy as A.I. grows more sophisticated and capable of performing human tasks.
The company is explicit that these are not forecasts, and the paper states: “The scenarios are not predictions and we attach no probabilities to them.”
In a benign scenario, A.I. turns out to be a minor technology and U.S. economic growth in 2030 is 1.6% above where it would be without the technology. Unemployment barely changes.
However, in a worst-case scenario, U.S. economic growth rises to 15.4% in 2030 and the economy begins to double every four years.
In such a scenario employment collapses by 21.5%, the unemployment rate among white-collar workers rises to 17.9%, and joblessness across the whole workforce hits 11.9%.
Additionally, office wages fall 11.5% and labour's share of national income drops from 60% to 45%, with capital income rising more than 80%.
Essentially, the machines would make the economy vastly richer while shifting the gains from workers to asset owners or people who own stocks and other A.I.-related investments.
The stark outlook was released days after Evan Hubinger, a safety researcher at Anthropic, warned that A.I. is advancing so fast there is a greater than 10% chance it “could kill all humans” within the next decade.
Anthropic is a privately held company but is expected to hold its initial public offering (IPO) this October at a valuation that could reach $2 trillion U.S.
Tech Insider
