Artificial intelligence capable of surpassing human capabilities. This concept, called “superintelligence”, is today more science fiction than reality, but it seems to be the basis of all recent efforts and is of interest to several companies, starting with OpenAI. For his boss, Sam Altman, this milestone could be reached within five years, as he explained in September during an event for the Axel Springer media group. However, many people do not want such an AI to see the light of day.
This Wednesday, October 22, more than 700 scientists, political figures, technology entrepreneurs and celebrities asked that this work be stopped due to the risks it would represent for humanity.
“We ask that the development of superintelligence be stopped as long as there is no scientific consensus on its possibility of building it in a controlled and safe manner and as long as it does not have the support of the population,” indicates the page of the initiative launched by the Future of Life Institute, a non-profit organization based in the United States that periodically warns against the harms of AI.
A potential danger to humanity
Among the signatories are several fathers of modern AI, such as Geoffrey Hinton, Nobel Prize winner in Physics in 2024, Stuart Russell, professor of computer science at the University of California at Berkeley, and Yoshua Bengio, professor at the University of Montreal.
The list also includes technology figures such as Richard Branson, founder of Virgin Group, and Steve Wozniak, co-founder of Apple; political figures such as Steve Bannon, former advisor to US President Donald Trump, and Susan Rice, national security advisor under President Barack Obama; religious leaders such as Paolo Benanti, advisor to the Pope and the Vatican’s leading expert on AI, but also celebrities such as the American singer will.i.am and Prince Harry and his wife Meghan Markle.
All consider that superintelligence poses risks to humanity such as “loss of freedoms, civil liberties, dignity and control.” According to them, such AI could also be dangerous to national security and even lead to the extinction of humanity.
“We can be in favor of creating more powerful artificial intelligence tools, to cure cancer, for example, and at the same time be against superintelligence,” he added.
Americans also seem to be worried about this. According to a survey published by the Future of Life Institute, 64% believe that superhuman AI should not be developed until it has been proven safe and controllable, or that it should never be developed. 73% also want strict regulation of advanced AI.
This petition, which already has more than 900 signatures, echoes a letter from researchers and executives in the AI sector, published a month ago during the United Nations General Assembly, which called for the establishment of “international agreements on red lines for AI” to avoid devastating effects for humanity.
Source: BFM TV
