A senior Anthropic safety researcher has warned there is a greater than 10 percent chance that artificial intelligence could kill all humans by the end of the decade, adding to a wave of alarming statements from industry insiders about the risks of superintelligent systems.
Key facts
- An Anthropic safety researcher said there is more than a 10 percent chance AI 'could kill all humans' by the end of the decade, according to The Verge.
- The warning came hours after a colleague resigned over fears AI labs are racing to build 'superhuman systems' they cannot control.
- The Guardian reported similar warnings were issued on both sides of the Atlantic over a 48-hour period.
- One warning raised the prospect of a 'Chornobyl-sized catastrophe' and asked whether AI poses a greater risk than nuclear weapons.
A senior safety researcher at the AI company Anthropic has warned there is a greater than 10 percent chance that artificial intelligence “could kill all humans” by the end of the decade, according to The Verge. The statement adds to mounting concern from within the industry about the pace at which powerful AI systems are being developed.
The warning came just hours after a colleague resigned over fears that Anthropic and its rivals are carelessly racing to build “superhuman systems” they cannot control, The Verge reported. The comments were made in a public post.
The Guardian reported that a series of alarming warnings from industry insiders emerged over a 48-hour period on both sides of the Atlantic, increasing pressure for curbs on artificial superintelligence.
Among the questions raised in those warnings, according to The Guardian, was whether artificial superintelligence poses a risk greater than nuclear weapons, and whether there is a significant chance of “a Chornobyl-sized catastrophe.”
The Guardian noted that these concerns relate to a technology that most people still think of as little more than “a more talkative search engine,” highlighting the gap between public perception and the fears being voiced by some experts.
The reports come amid broader debate over how seriously the public and policymakers should take such warnings, with The Guardian framing its coverage around the question of how much weight to give to the statements of industry figures.
Why it matters
Warnings from insiders at leading AI labs carry weight because these are the people building the technology, and their concerns are fueling calls for regulation of advanced AI. For a general reader, the debate highlights a growing tension between rapid commercial development of AI and questions about whether the risks can be safely managed.