Recently, more than 350 AI industry experts jointly issued a short statement of only 22 words, warning that the possible existential threat of AI to mankind should become a global priority issue like epidemics and nuclear war. The statement was posted on the website of the nonprofit Center for Artificial Intelligence Safety. While AI experts, journalists, policymakers and the public are increasingly discussing the broad risks posed by AI, it remains difficult to voice concerns about the most serious risks of advanced AI, the center said.
This succinct statement aims to overcome this obstacle, spark discussion, and provide consensus among a growing number of experts and the public who are serious about the serious risks of AI. The statement was signed by more than 350 AI practitioners, including Google DeepMind CEO Demis Hassabis, OpenAI CEO Sam Altman, and AI "godfathers" Jeffrey Hinton and Joshua Bensio. Hinton and Bencio won the 2018 Turing Award for their AI work, but one of the winners, Facebook AI chief scientist Yang Likun, was unsigned.
Some of the signatories have previously signed an open letter calling for a "moratorium" on the development of AI more advanced than GPT-4 for at least 6 months. The letter was criticized, with some experts arguing that the AI risks were exaggerated, and others agreeing that the risks existed but disagreed with the measures in the letter. Dan Hendrix, executive director of the Center for AI Security, told The New York Times that the May 30 statement was very concise and did not propose any specific ways to mitigate the AI threat to avoid such disagreements. "We don't want to come up with a detailed menu of 30 potential interventions," Hendrix said. ”
In summary, AI experts issued a brief joint statement warning that the possible existential threats posed by AI deserve broader attention. The statement is concise and aims to build consensus on AI risks and provide a basis for further discussion and countermeasures.