If deep down the USA and China also cooperate in AI. Report Ft
Despite economic competition and political conflict, America and China are still collaborating on the regulation of AI. What the Financial Times revelations say
A scoop from the Financial Times revealed how in the prodigious and delicate field of Artificial Intelligence (AI) an embryo of cooperation in security matters is developing between the two same American and Chinese superpowers that compete fiercely in the field of advanced technologies. Here is what the newspaper revealed about the two confidential meetings held in Geneva with the participation of experts and scientists from the two countries and the approval of their respective governments.
Close Encounters on AI in Geneva
According to Financial Times sources, between July and October last year at least two meetings were held in Geneva in which scientists and experts from OpenAI, Anthropic and Cohere discussed the risks with some representatives of Tsinghua University and other Chinese state institutions. resulting from AI.
What was discussed?
Those who participated in meetings organized by the Shaikh Group and approved by the governments of Washington and Beijing as well as that of London told the newspaper that the primary objective of the meetings was to find a safe scientific approach to develop even more sophisticated AI applications.
During the talks, the American and Chinese experts would have discussed the areas in which technical cooperation is imaginable and desirable and would have even gone so far as to explore policy ideas on AI which were reflected in the discussions held in the two most important international events on the topic occurred last year, the one convened by the UN Security Council in July and the summit organized by the British government in November.
The words of the head of the Shaikh Group
Confirming the revelation, the chief executive of the Shaikh Group – a private organization created to promote dialogue between actors at risk of conflict – explained to the FT that "we took the opportunity to bring together the key actors from the USA and China who work on the AI”.
“Our aim – specifies Salman Shaikh – was to highlight the vulnerabilities, risks and opportunities underlying the development of AI models that are shared across the planet. Recognizing this fact can, we believe, lay the foundation for collaborative scientific work that leads to the development of global safety standards."
Chinese confirmation
This embryonic cooperation found indirect words of confirmation in a statement from the Chinese Embassy in Great Britain, according to which “China supports efforts to discuss AI governance and to develop the necessary frameworks, rules and standards on the basis of a broad consensus."
Beijing, the statement adds, “stands ready to carry out communications, exchanges and practical cooperation with various parties on global AI governance, and ensure that AI develops in a way that advances human civilization.”
The meaning of the meetings according to the Ft
The City newspaper underlines how the Geneva talks represented a rare sign of Sino-American cooperation in the same field of cutting-edge technologies which sees the two superpowers engaged in a head-on and no-holds-barred clash, as demonstrated by the US ban on exports of the most advanced chips produced by companies such as Nvidia which are needed to develop sophisticated AI software.
The potential existential risks for humanity posed by AI and reported by many parties are therefore pushing two absolute protagonists of this innovation to develop in a concerted manner security protocols considered of common interest.
The call on AI management
Among other things, the joint declaration signed in November by Chinese scientists and Western academics involved in the development of AI is recalled, in which they warned of the "existential risk to humanity" posed by advanced AI models, calling for their strict regulation.
There was also the name of the most famous Chinese computer scientist, Andrew Yao, at the bottom of that appeal which asked, among other things, for the creation of an international authority, the inclusion of instant "shutdown" procedures and the obligation for developers to allocate at least 30% of their research budget into AI safety studies.
This is a machine translation from Italian language of a post published on Start Magazine at the URL https://www.startmag.it/mondo/cooperazione-stati-uniti-cina-ia/ on Sun, 14 Jan 2024 06:46:01 +0000.