Abstract:
On September 28, according to Bloomberg, Microsoft co-founder Bill Gates said that the US government’s protective measures against the potential catastrophic risks of AI will not hinder the United States from competing with China. This view contrasts with the largely hands-off approach of US President Donald Trump.
Gates
In an interview with NBC's "Meet the Press," Gates said: "Of course, all countries should want to establish these protective measures. So, if someone says that China will not take these protective measures, I completely disagree with that statement."
With the rapid development of AI technology, the debate surrounding AI safety is heating up. Gates is one of many people in the technology community who has expressed his views. Some warn that AI could put humans at risk. Gates said in an interview published by NBC last week that AI is a powerful enough tool that it may even "lead to the death of a billion people."
Gates said that AI needs not only self-regulation, but also stronger protective measures. He called on the U.S. Congress to pass relevant legislation. "Law enforcement and politicians need to be involved in discussions about how protective measures and surveillance mechanisms should be designed," he said, "and this must become a hard requirement."
Facing sluggish approval ratings and with Republicans fighting to retain control of Congress in the November election, Trump has described warnings about the existential risks posed by AI as a "hoax," saying they threaten U.S. innovation and allow China to profit from them.
When asked about Trump's remarks, Gates said: "This is not a hoax at all."
Gates said he expects China to join the ranks of supporting AI security because China does not want to be a target of attacks. He also believes Trump may change his position "after hearing from me and others that building safeguards will not prevent us from dealing with what he sees as nation-state competition."
He previously stated in an article published in August that criminals may use increasingly powerful AI tools to launch cyber attacks on hospitals, financial institutions, power grids and government systems, commit fraud, spread false information, and carry out other forms of large-scale damage.
In an interview with NBC, Gates said that the so-called "emergency stop switch" is not enough and cannot replace supervision at an earlier stage. The emergency stop switch is designed to ensure that humans still have the ability to stop or limit powerful AI models when they behave dangerously.
"We do need to monitor any advanced AI model and accurately record everything it performs. The United States should ensure that this is done. China will also want to do the same." Gates said.
Comments