At the risk of being apocalyptic, it was hard not to be stunned by a 2022 survey that found just under half of artificial intelligence researchers polled thought that once the technology had matured, there was at least a 10 per cent chance it would end in an “extremely bad outcome” such as “human extinction”.
The AI researchers were among the top in their field. In a previous separate survey, most thought that this tipping point – where AI is “able to accomplish every task better and more cheaply than human workers” – would most likely occur within the next 40 years. Some thought it would be much, much sooner.