According to a report on Wednesday local time in Reuters, White House officials confirmed that the CEOs of Google, Microsoft, OpenAI and Anthropic, an artificial intelligence (AI) research company, will meet with US Vice President Kamala Harris and senior government officials on Thursday local time to discuss key issues of AI.
A White House official revealed that Harris and government officials will tell these corporate executives involved in the AI field at the meeting that they have the responsibility to reduce the potential harm of artificial intelligence tools. In the invitation letter, US President Biden also hoped that technology companies must ensure the safety of their products before offering them to the public.
Last month, Biden emphasized the necessity of supervising products related to artificial intelligence, although he said whether artificial intelligence poses a real threat remains to be seen. Biden said that social media has explained the harm that powerful AI technology may cause without proper protection measures.
In fact, since the advent of ChatGPT last year, it has caused great social controversy. Since the capabilities of AI have far exceeded people's expectations, critics' concerns about the rapidly developing AI technology include invasion of privacy, prejudice, fraud and spreading rumors and false information. Industry technology leaders have also recently called for vigilance against the risks brought by AI.
On March 29th, thousands of scientists and technicians jointly published an open letter, calling on all artificial intelligence laboratories to immediately suspend the training of systems more powerful than GPT-4 until the related risks and problems are solved. This appeal has received a lot of attention and response. In just two days, more than 2,300 people have signed it.
Geoffrey Hinton, the "godfather of AI" who recently announced his departure from Google, said that he was worried that the Internet would be full of fake photos, videos and words, and ordinary people "could no longer tell the truth from the false". In addition, Hinton is also worried that the "future version" of technology will pose a threat to human beings, because "they often learn unexpected behaviors from a large amount of data analyzed" and it is impossible to imagine how to prevent bad people from using AI to do bad things.
Hinton said that artificial intelligence will not only lead to the disappearance of a large number of jobs, but also make people unable to know what is real. If artificial intelligence becomes much smarter than human beings, it will manipulate human beings; When artificial intelligence knows how to program, it will try to get around the restrictions set by human beings.
In fact, the threat is already obvious. For example, IBM will replace more than 7,000 jobs with artificial intelligence. Krishna, CEO of IBM, said on May 1 that the recruitment of internal functional departments such as human resources of the company will be suspended or postponed, and there are about 26,000 positions that are not directly oriented to customers in IBM. In the next five years, 30% of the jobs will be replaced by artificial intelligence and automation. This means that about 7,800 people will lose their jobs.
At present, governments all over the world are actively seeking regulatory measures to catch up with the pace of technological progress. In the United States, the Ministry of Commerce officially publicly solicited opinions last month, asking all sectors to comment on the accountability measures of artificial intelligence. In addition to the United States, EU legislators are also accelerating the supervision of artificial intelligence technology.
Last week, the European Union proposed new draft legislation, requiring developers of artificial intelligence tools such as ChatGPT to disclose the copyright materials they used when building AI systems. As a regulatory scheme, Vestager, executive vice-president and competition commissioner of the European Commission, put forward a scheme to display labels such as "made with AI" on articles and images generated by AI, and also revealed the idea that enterprises that provide generative AI are obliged to explain.
EU legislators are still discussing how to introduce the EU artificial intelligence bill, and generative AI(AIGC) has become one of the most concerned contents. The European Commission will vote on the relevant agreement on May 11th. If it is passed, it will enter the next stage of tripartite negotiations, when EU member States will debate the contents of the negotiations with the European Commission and the European Parliament.
Yang Haiping, a researcher at the Securities and Futures Research Institute of the Central University of Finance and Economics, said that if the generative artificial intelligence is not regulated, there may be risks of data security and network security, as well as impacts on production order, life order and social order caused by misuse and abuse of artificial intelligence. More importantly, because the generative artificial intelligence generates text, pictures, sounds, videos, codes and other contents based on algorithms, models and rules, if the responsibility of providers is not strengthened and strict supervision is not implemented, technological innovation may break through the bottom line of scientific and technological ethics, deviate from the direction of scientific and technological goodness, and cause incalculable and irreparable losses.
Yang Haiping said that foreign scientists called for the suspension of the development of more powerful AI systems, and China introduced new regulations on the service management of generative artificial intelligence, all at the critical moment when artificial intelligence made a key breakthrough, and rational preventive measures were taken against possible risks of artificial intelligence. The supervision of artificial intelligence requires the international community to assess and draw a unified and clear safety bottom line for all countries to abide by.