Senate Judiciary Hearing: AI is difficult to control and may be used by malicious elements to develop biochemical weapons

On July 26, according to foreign media TechCrunch, the U.S. Senate Judiciary Committee held a hearing on artificial intelligence yesterday. During the meeting, many AI industry practitioners warned , Human beings cannot properly control the rapidly developing AI, so AI may cause serious harm in the next few years, and malicious elements may use AI tools that bypass supervision to create biochemical weapons.

Richard Blumenthal, chairman of the Senate Judiciary Committee subcommittee, who chaired the hearing, said that humans have proven themselves capable of inventing new technologies that were unbelievable at the time. He compared AI to America's second "Manhattan Project" or NASA's second "Manned Moon Landing Plan," but the industry representatives in the audience showed more concerns about AI. Dario Amodei, CEO of the American AI startup Anthropic, said at the hearing that

he is worried that cutting-edge AI technology may be used to create dangerous viruses and other biochemical weapons in just two years.

, the company has previously received an investment of US$300 million (IT House note: approximately 2.147 billion yuan) from Google, and relies on Google's data center to train "more reliable" AI models. Yoshua Bengio, an AI professor at the University of Montreal and the "father of modern AI science", said that he was shocked by the huge leaps achieved by models such as GPT-4. He believed that

currently countries should actively carry out Cooperate to control the development of AI and establish a system similar to nuclear technology regulatory rules

. Stuart Russell, a professor of computer science at the University of California, Berkeley, believes that the current way artificial intelligence works is full of unpredictability. Compared with other controllable and powerful technologies, artificial intelligence at this stage lacks a It is a super management tool, so we cannot accurately understand the true intentions of artificial intelligence and all the actions it may take

According to reports, in the past six months, the industry has generally believed that the point in time when "super AI" may appear in Hollywood blockbusters It has been advanced from decades to just a few years, which has triggered widespread concerns from all walks of life about the various problems that "super AI" may bring

The above is the detailed content of Senate Judiciary Hearing: AI is difficult to control and may be used by malicious elements to develop biochemical weapons. For more information, please follow other related articles on the PHP Chinese website!