Geoffrey Hinton, a computer scientist renowned for his work on artificial neural networks and often referred to as the 'godfather of AI,' has issued a stark warning about the potential dangers of superintelligent artificial intelligence. He contends that the development of AI systems surpassing human intelligence could lead to human extinction.
Hinton stated that it would be "very foolish" to pursue superintelligence now, given the lack of a scientific consensus on how to develop and control such systems safely. "Losing control over AI smarter than ourselves could be catastrophic and could even lead to human extinction," he explained.
His comments coincide with a proposed bill in the UK Parliament, the Artificial Superintelligence Security Bill. Drafted by ControlAI and put forward by Labour MP Alex Sobel, the bill seeks to prohibit the development of superintelligent AI within the United Kingdom. According to ControlAI, the campaign has garnered support from over 135 cross-party UK parliamentarians. The organization asserts that while specialized AIs can enhance science, medicine, and public services, superintelligent systems pose a threat to national and global security.
This push for regulation contrasts with the views of some industry leaders. Sam Altman, CEO of OpenAI, has previously advocated for more regulation but now expresses concern about potential "overregulation." Altman has argued that overly strict regulations could hinder innovation and that it is impossible to establish safeguards before problems arise. Hinton, however, believes this stance could lead to humanity's downfall, stating last year, "If we can't figure out a solution to how we can still be around when they're much smarter than us and much more powerful than us, we'll be toast." He emphasized the need for a counterbalance to those in the tech industry who oppose AI regulation.
Superintelligence is defined as the point at which an artificial system's capabilities exceed those of any individual, company, or nation. ControlAI suggests that the UK can balance the benefits of AI with its risks by implementing binding regulations on the most powerful AI systems.