{"text":[[{"start":6.96,"text":"OpenAI has called on the Trump administration to lead an international coalition setting standards for AI systems ahead of an appearance at the UN Security Council amid rising fears about the risks of the technology."}],[{"start":20.42,"text":"The ChatGPT maker laid out proposals on Monday for global co-ordination to evaluate the capabilities and risks of AI models, particularly as they develop the ability to train themselves, echoing calls by rivals including Anthropic and Google for more robust standards and greater co-operation between nations."}],[{"start":39.96,"text":"The proposals preview the pitch OpenAI chief executive Sam Altman will make when he briefs the UN Security Council on Wednesday and attends a White House state dinner with Chinese President Xi Jinping and US President Donald Trump on Thursday."}],[{"start":54.4,"text":"Treasury secretary Scott Bessent on Sunday told the FT the US and China had agreed to set up an AI dialogue after talks with Chinese vice-premier He Lifeng in New York."}],[{"start":65.72,"text":"OpenAI also called for the establishment of secure channels between the two nations leading the AI race to share emerging threats. “Dialogue between the United States and China on AI safety and security would be a positive step,” said the company."}],[{"start":80.28,"text":"The proposal comes amid growing alarm about the potential of AI. Both Anthropic and OpenAI have reported multiple incidents of their models hacking into external systems, a phenomenon that has become increasingly concerning as the capabilities of AI tools rapidly advance."}],[{"start":95.6,"text":"Those improvements have also prompted warnings from researchers within the companies about the potential of rogue or misused AI systems wiping out humanity."}],[{"start":105.16,"text":"Anthropic chief executive Dario Amodei, Altman and SpaceX boss Elon Musk called for a slowdown in AI development earlier this month — a rare sign of consensus in the highly competitive sector."}],[{"start":116.4,"text":"However, Trump has rejected calls for a slowdown, pointing to competition from China. The president has said that “whoever wins AI, wins”."}],[{"start":124.96,"text":"David Sacks, an influential figure in Washington who has advised Trump on AI, was also critical of industry pleas for government intervention: “The easiest way not to build superintelligence is for you to agree not to build it. Demanding your preferred regulatory framework as the price of that will look like blackmail of the public and the political system,” he wrote on social media."}],[{"start":144.72,"text":"OpenAI’s new proposals are in line with a call from Demis Hassabis, co-founder and chair of Google DeepMind, for a “US-initiated effort [to] provide a strong starting point for creating shared international standards on Frontier AI” in July."}],[{"start":160,"text":"Hassabis outlined a new public-private body, “much like the Financial Industry Regulatory Authority (Finra), with a board that includes independent leading technical experts and open-source representatives”."}],[{"start":172,"text":"He said something similar to the international financial regulatory body was necessary to manage a technology whose impact would be “perhaps 10x of the Industrial Revolution at 10x the speed”."}],[{"start":183.84,"text":"On Monday, OpenAI proposed establishing international standards for evaluating AI models via the existing network of national AI safety institutes."}],[{"start":193.52,"text":"“This approach can draw lessons from areas such as aviation and financial stability, where countries have developed common technical standards and trusted channels for co-operation without giving up national authority,” said the company."}],[{"start":205.84,"text":"In particular, international co-operation would be vital in managing recursive self-improvement, where more of the work of training AI models is done by other AI systems."}],[{"start":215.92,"text":"“Done without appropriate care and caution, RSI could result in humans losing practical control over AI development, unable to provide oversight on research processes they no longer understand,” said the company on Monday."}],[{"start":232.37,"text":""}]],"url":"https://audio.ftcn.net.cn/album/a_1790037632_9170.mp3"}