Welcome to the loop Time The new information ballot about AI twice a week. We publish these publications both stories at Time.com, as well as emails. It was a tense week for our team: Tarin Pilley was in place during the UN General Assembly in New York, while Harry Butita and Nikita Ostrovsky were at the “All in AI” event in Montreal.
If you read this in your browser, why not subscribe to the next one to be delivered directly to your mailbox?
What to know: the UN takes for art
AI takes the podium – The General Assembly of the United Nations met this week in New York. While members of the Assembly spent most of their time on crises in Palestine and Sudan, they also devoted a good part of AI. On Monday, the Laureate of the Nobel Prize of the World, Maria Ressa, drew attention to the campaign for the “red lines of AI”, begging the governments to get together to “prevent everywhere unacceptable risks” from AI. More than 200 outstanding politicians and scientists, including 10 Nobel Prize winners, Signed on statement.
“New curtain” – On Wednesday, the Security Council participated in open debate about “artificial intelligence and international world and security.” For more than three hours, each country takes turns taking turns, delivering approximately the same width: AI held the promise for both good and harm. Again and again, representatives stated that AI is not science fiction, but by the fact of modern life, and that international regulatory fences should have been developed immediately, especially around autonomous weapons and nuclear.
One of the most interesting prospects came from Belarus, which attracted attention to the growing global inequality in the development of AI. “This time a new curtain is created, not ideological, but technological … for the separation of the West and other part of the world, as well as attract the global countries of the majority to the era of neocolonialism,” the representative said. “This leads to a dead end and the abyss. AI should be available and available for all countries that are not with the exception. ”
New institutions – This section was served by Piley’s tarin from the UN headquarters. On Thursday, a three-hour “informal meeting” took place at the UN, which began the global dialogue for managing an artificial intelligence-new forum for all UN Member States, as well as companies in the private sector and civil society organizations in order to coordinate the management of AI. “For the first time, each country will take place at the AI table,” said the UN Secretary General Antonio Guterierires. Guterres also announced that the candidates were open to candidates to join the new International Independent Scientific Group on AI, intended to provide impartial scientific data on the influence of technology. “Soon I will begin consulting with member states, potential sponsors and partners to create a global fund for the development of the potential of AI,” he added.
This was followed by Bhutan, the UAE and China, sharing reports on how AI supports economic growth, and many “interventions” by government ministers, technical leaders and representatives of the academic and civil society – many of them, such as the Minister of Nigeria Bosun Tidjani, Godfitsa Aioshua Bengii, and at present, the host, in general. Ordinary calls came so that the spaces were prepared, built, benefits were obtained, interested parties for consulting and inequalities for compensation. “The growth of AI does not remain unchanged,” said Pedro Sanchez, Prime Minister -Minister of Spain. “But this cannot be uncontrollable.”
Does anyone listen? It is unclear whether something from this real influence on the development of AI will be, since the company of Silicon Valley is not related to the recommendations of the UN. But the compilation of the UN is really a more holistic and global inclusive path forward for this technology.
Tracking SB 53
Take two – Last year, the Governor of California Gavin News imposed a veto on the state bill on the security of AI, SB 1047 and told the researchers to come up with an alternative. They didAnd the lawmakers accepted their recommendations and accepted the updated, unwashed version, SB 53. Once again, all eyes turned to Newsom: he really will remain true to his word and supports this version, the cave of the technological industry, that the legislation has gone too far?
Widespread support – While Newsom has not yet put the pen on the paper, he gave an idea of his support from the bill on the stage in New York on Wednesday. “We have an account – for me, this is on my table – that we think that it comes to the right balance,” he saidThe field “We worked with the industry, but we did not obey industry.”
SB 53 is important because it causes protection against informants for AI employees and requires that the largest developers publish security plans and report security incidents. Anthropic supports this, as well as Dean Ballwho worked on the Trump action plan.
A gradual victory – And the security groups of AI note the bill as a major victory, despite the fact that it lacks many of the most strict protection that 1047 had last year. “The image of politics necessarily concerns a compromise. You want to advance the envelope as far as possible, but you cannot abandon the incrementalism, ”says Sasha Havort, executive director of the technology supervision project. “The protection of the informants is huge, and the fact that the rules in SB 53 apply to the largest AI developers are crucial.”
Time in action
One of the AI Red Lines campaign was Yoshua Bengii, who, as it would be fate, had an interview on the stage at Harry Boet at the All in Montreal conference on Wednesday. (The conference is not related to the popular technical podcast with the same name.) The couple spoke about the development of models of reasoning, that the government is turning away from the security and its new non -profit organization Lawzero, which is aimed at changing the security of AI in the face of commercial imperatives.
“It's not that these systems are going to kill someone else tomorrow,” he said to Harry. “But future generations of these systems, if science continues to advance, will have more powerful and stronger reasoning abilities, more and more knowledge.”
“And if we cannot make sure that they act in accordance with our norms, then they can be used by unhappy people for immoral things, and we could lose control of them,” he continued. “Therefore, even if we do not have confidence in these capabilities, the bets are so high that we need absolutely to make sure that an accident with a low probability will not even happen.”
You can see all the interviews HereField
What are we watching
I just watched the movie Soundtrack to the coupWhich was nominated for the best documentary in Oscars this year. Although the film is not about AI, I discovered a lot of resonans between his history about the Cold War at this moment AI, especially in his description of two global superpowers fighting for control of resources and Mindshare.
The film is focused on the Congo, which was strategically important for both the United States and the USSR for its uranium mines, which were necessary to create atomic bombs. In the film, shots are shown by the Belgian Prime Minister Gaston Escens, who says that the occupation of the Congo is “not to satisfy colonial or imperial aspirations, but to fulfill the mission of civilization for the benefit of less developed people … for its salvation and ascension.” This language sounds Terribly familiar Technical leaders specialize for rhetoric, especially when they emphasize that the development of AI is as important for national security as nuclear energy.
Since technical leaders deploy strategies and narratives that repeat the countries of colonialism (Read Karen Hao for more), Congo is again in the crosshair, and miners collect cobalt, which nourishes smartphones, computers and electric cars in The working conditionsField
You can rent Soundtrack to the coup For $ 4 On YouTubeField
With reporting Thin Piri / New York.
