IA: China, USA and Brazil sign agreement for joint action – 11/01/2023 – Tech

China agreed this Wednesday (1st) to work with the United States, the European Union and other countries to collectively manage the risks posed by artificial intelligence (AI), with the aim of charting a safe path for the technology.

Some technology company executives and political leaders have warned that the rapid development of AI poses an existential threat to the world if left unchecked, prompting a rush by governments and international institutions to create safeguards and regulations.

In a first for Western efforts to manage the safe development of AI, a Chinese vice minister joined US and EU leaders and technology company presidents including Elon Musk of X (formerly Twitter), and Sam Altman from ChatGPT at Bletchley Park, home of WWII codebreakers in the UK.

In addition to China and the USA, Brazil and 25 other countries signed the “Bletchley Declaration”, stating that countries need to work together to establish a common approach to supervision. Brazil and Chile were the only Latin American countries with representatives at the summit.

The declaration set out a dual agenda, focused on identifying risks of shared concern and building scientific understanding about them, as well as developing policies between countries to mitigate them.

Wu Zhaohui, China’s vice minister of science and technology, said at the opening session of the two-day summit that Beijing was ready to increase collaboration on AI security to help build an international “governance structure.”

“I don’t know what the fair rules necessarily are, but you have to start with understanding before you do oversight,” Musk told reporters. He added that a “third-party referee” could be used to alert when risks arise.

While the European Union has focused oversight over AI on data privacy and surveillance and their potential impact on human rights, the British summit is looking at the so-called existential risks of highly capable general-purpose models called “frontier AI.”

Mustafa Suleyman, co-founder of Google Deepmind, said he doesn’t think current frontier AI models pose any “significant catastrophic harm,” but said it makes sense to plan ahead as the industry trains ever larger models.

The US made it clear on Tuesday (31) that the invitation to Beijing largely came from the United Kingdom. British Ambassador Jane Hartley told Reuters that “this is the UK’s invitation, not the US’s.”

US Vice President Kamala Harris’ decision to give a speech in London on Wednesday about her government’s response to AI and to hold some meetings with participants outside the summit means the US could exit the group in advance, also raised some suspicions.

But British representatives denied this, stating that the interest was in obtaining as many voices as possible.

Just days after US President Joe Biden signed an executive order on AI, his administration used the British summit to announce that it will launch an “AI Safety Institute.”