Singapore’s Vision for AI Safety Bridges the US-China Divide

Government Singapore released A blueprint After the meeting of AI researchers from the United States, China and Europe, today for global cooperation on the protection of artificial intelligence. This document offers a joint vision to work on AI’s security through international cooperation rather than competition.

“Singapore is one of the few countries on the planet that combines with both the East and the West,” says Max Tagmark, a MIT scientist. “They know that they are not going to make themselves (artificial general intelligence) – they have done this with them – so there is a lot of interest in their interests that the countries that are going to build it will talk to each other.”

These countries believed that the most likely AGI was likely to be constructed, of course. , America and China – and still they feel more intended to improve each other besides working together. In January, after the Chinese startupdipidic releasing a modern model, President Trump called it a “wake -up call for our industries” and said the United States “needed to be laser focused on winning”.

Singapore consensus on the priorities of global AI seafy research calls researchers to cooperate in three key fields: studying the dangers of Frontier AI models, finding safe ways to build these models, and developing methods to control the latest AI system.

The consensus was prepared on April 26 this year at a meeting held on April 26 with the International Conference (ICLR) of a Premier AI event held in Singapore.

Researchers from Openi, Anthropic, Google Deep Mind, Z, and Meta attended the AI ​​Safety event, such as MIT, Stanford, Singhwa, and the Chinese Academy of Sciences. Experts at the AI ​​Safety Institute also participated in the United States, the United Kingdom, France, Canada, China, China, Japan and Korea.

“In the era of geographical political pieces, this comprehensive synthesis of modern research on AI seafy is a passionate symbol that the global community is combined with a joint commitment to form a safe future,” said Zio Lane, Dean of Singhwa University, in a statement.

The development of rapidly capable AI models, some of which has amazing abilities, have caused researchers to worry about many risks. While some focus on near -term disadvantages, including the problems arising caused by the prejudiced AI system or its possibilities Guilty to bringing technology to workA significant number of AI believes that AI can pose an existential threat to humanity as it has begun to wander humans in more domains. These researchers, sometimes called “AI domestic”, fear that models can deceive and manipulate humans to achieve their goals.

AI’s capacity has also talked about arms race between the United States, China and other powerful countries. This technology is considered important for economic prosperity and military dominance in policy circles, and many governments have tried to put their own views and regulations into practice in which it should be developed.