Microsoft’s chief scientist warned that the ban proposed by Donald Trump on railing at state level on artificial intelligence will slow the development of border technology rather than accelerate.
Dr. Eric Horvitz, a former technological advisor to Joe Biden, said that regulatory prohibitions “we will” hold us “and” could be in contradiction to make good progress not only to advance science, but to translate it into practice “.
The Trump administration has proposed a 10 -year ban on American states creating “any law or regulation limiting, restricting or regulating artificial intelligence models, artificial intelligence systems or automated decision -making systems”.
It is partly motivated by the fears of the White House, China could otherwise win the AI race at the human level, but also the pressure of technological investors, like Andreessen Horowitz, a first investor in Facebook, which arguments Consumer uses should be regulated rather than research efforts. His co-founder, the donor of Trump, Marc Andreessen, said earlier this month that the United States was in a two-horsepower race for AI supremacy with China. The American vice-president, JD Vance, recently said: “If we take a break, (China) does not take a break? Then we find ourselves … enslaved at the mediated AI (in China). “
Horvitz said he was already concerned that “AI was exploited for inappropriate disinformation and persuasion” and for its use “for malicious activities, for example, in the biological risk space”.
Horvitz’s pro-regulation comments came despite the information according to which Microsoft Be part of a lobbying of Silicon Valley lobbying with Google, Meta and Amazon, to support the ban on individual American states regulating AI for the next decade which is included in the Trump budget bill which passes through the Congress.
Microsoft is part of a lobbying campaign to urge the US Senate to adopt a decade moratorium on individual states with their own efforts to legislate, the Financial Times reported last week. The ban was entered in Trump’s “Big Beautiful Bill” which he wishes to adopt by independence day on July 4.
Horvitz spoke at a meeting of The association for the progress of artificial intelligence On Monday, when he said: “It is up to us as scientists to communicate to government agencies, in particular those at the moment which could make declarations on any regulations, (that) this will retain us.
“Advice, regulations … reliability controls are part of the progress of the field, which makes the field faster in many ways.”
Speaking during the same seminar, Stuart Russell, IT professor at the University of California in Berkeley, said: “Why do we deliberately allow the release of a technology which, even as its creators, have a risk of 10% to 30% … to cause human extinction? We will never accept anything about this level of risk for any other technology. ”
After promoting the newsletter
The apparent contradiction between Microsoft’s chief scientist and the company’s lobbying effort reports intervenes in increasing fears that the non -regulated development of AI could present catastrophic risks for humanity and is motivated by private priorities in terms of short -term priority.
Microsoft has invested $ 14 billion (10 billion pounds Sterling) in Openai, the Chatgpt developer, including CEO Sam Altman who predicted this week: “In five or 10 years, we will have large human robots and they will simply get down to the street by doing things … I think it would be one of the most amazing moments.”
The predictions of the moment when the general artificial intelligence of the level of man (AG) will be reached varies from a few years to decades. Meta chief scientist Yann Lecun said AGE could be in decades, while last week his boss, Mark Zuckerberg, announced an investment of $ 15 billion in order to achieve “superintendent”.
Microsoft refused to comment.