自由1688

Huang Renxun Again Refutes AI Doomsday Claims: It Won’t Wipe Out Humanity; the Probability of a World End in 2030 Is 0

On September 19, according to a report by Bloomberg, NVIDIA CEO Jensen Huang again refuted AI doomsday theories, saying that this technology does not pose an extinction risk to humans. Huang was interviewed by CBS News. When the host asked whether he agreed with President Trump’s claim that “AI concerns” are a “scam,” Huang seemed reluctant to answer directly, but he clearly stated that he disagrees with the prediction that AI will destroy humanity in 2030.

How powerful will AI be three months from now? Even top researchers don’t dare to predict.

A researcher involved in work on the Millennium Prize Problems used to be willing to forecast AI progress over the next 12 months, but now won’t even venture to judge what will happen three months from now. On September 17, OpenAI researcher Noam Brown relayed this detail during an interview with Dwarkesh Patel. The conversation also covered multi-agent systems, AI’s involvement in manufacturing the next generation of AI, and increasingly difficult safety testing.

2030 will destroy humanity? Jensen Huang: Probability is 0

Huang Renxun refutes the “AI doomsday theory” again

@Leo - F0

Novo Nordisk teams up with Anthropic to accelerate new drug development

Danish pharmaceutical giant Novo Nordisk will collaborate with AI company Anthropic to use AI tools to advance drug research. Novo Nordisk said on Wednesday local time that the company will use Claude Science, a platform built by Anthropic for scientific research, to speed up the process of discovering and developing new drugs, with a focus on tackling scientific questions that both parties believe could have the greatest real-world impact.

Novo Nordisk is the manufacturer of diabetes and weight-loss drugs Ozempic and Wegovy. The partnership is another example of major pharmaceutical companies increasingly ramping up investments in AI in recent months. Several large drugmakers, including Eli Lilly, Merck & Co., and Roche, have been heavily betting on AI, hoping to discover and produce drugs faster with the help of this technology.

Intel CEO: Memory shortages will further worsen; power and cooling technologies are the key bottlenecks

Intel CEO Ren Xu? Chen Zhixuan? said that the memory shortage problem will further intensify. He also added that ensuring power supply and developing cooling technologies will become the core challenges for the semiconductor market. On the 15th, at the AI Infrastructure Forum held in Santa Clara, California, when asked about which industry problems he hopes founders of start-ups in the field of AI infrastructure will solve, Chen Zhixuan discussed the memory shortage issue.

Obama Rarely Intervenes in the AI Regulation Debate, Urging the U.S. to Pass Proactive Legislation to Avoid “Potential Catastrophe”

Former U.S. President Barack Obama said he supports the creation of federal laws and regulations to establish standards for AI safety, warning that if mishandled, the technology could lead to a “potential catastrophe.” In a statement Monday, Obama said AI policy “should be at the center of public debate,” adding that policymakers in Washington must “take proactive action,” develop relevant plans “to address serious safety concerns, anticipate the impact of AI on jobs and the next generation, and ensure that the benefits brought by AI are broadly shared.”

The former president criticized the approach to AI regulation advocated by President Donald Trump, saying it is one of laissez-faire. While he did not name his successor directly, he said that “voluntary standards set by a handful of technology companies are not enough.”

An executive order signed by Trump in June this year created a voluntary participation framework that allows AI developers to share frontier models with the government in advance so potential safety risks can be assessed.

Obama said: “Given what we know about technology, we can’t put AI back in the box. But we can decide together how it should be developed, how it should be used—not sit back and let AI and its consequences come crashing down on us.”

Obama’s rare public comments are his deepest remarks to date on AI. As leading figures in the industry and members of Congress have increasingly called for slowing the development of frontier models, AI has become a core issue in national discussions in the United States.

With a lack of clear leadership figures within the Democratic Party, many Democrats are still seeking Obama’s advice and guidance on policy and political strategy. By stepping into this debate, Obama is attempting to position the Democrats as the party that supports strengthening AI safety measures, in preparation for the midterm elections in November and the start of the 2028 presidential election cycle.

Anthropic’s CEO Says the “China Factor” Is the Biggest Challenge in Slowing the AI Race

Debate in the artificial intelligence industry over whether development speed should be slowed is heating up rapidly, and China has been pushed to the center of this discussion. Recently, Dario Amodei, CEO of Anthropic, publicly stated that if the world attempts to set speed limits on AI development, the biggest practical challenge would not lie in the technology itself, but in how to deal with the possibility that competitive players such as China may continue to accelerate their efforts.

NVIDIA Considers Investing $10 Billion in Anthropic’s IPO—Super IPO Valuation Could Reach $2 Trillion

AI company Anthropic is preparing for what could become the largest initial public offering (IPO) in global history, and NVIDIA is considering participating as a cornerstone investor. Insiders say Anthropic plans to raise as much as $100 billion through the IPO, with the company’s post-listing valuation potentially reaching about $2 trillion; NVIDIA is considering investing up to $10 billion to buy shares.

If the deal is ultimately completed at this scale, Anthropic’s listing would set a new record for the largest IPO ever, and it would also serve as a major test of the capital markets’ enormous valuation of cutting-edge AI companies and the huge demand for computing power. Relevant negotiations are still ongoing, and the specific investment amounts and deal terms may change. Because the discussions are private, insiders requested anonymity.

The news that NVIDIA might participate in Anthropic’s IPO was not previously made public. If the transaction is ultimately finalized, NVIDIA would become an important strategic backer at the time of Anthropic’s listing, further deepening the already very close business relationship between the two. As one of the world’s largest AI accelerator suppliers, NVIDIA is also one of Anthropic’s most important customers.

Samsung increases investment in Apple foldable iPhone Duo panel supply

Samsung Display has drawn up a new business plan, projecting a substantial increase in the number of foldable OLED panels it will supply to Apple next year, as well as OLED panels for iPads. Under the plan, the supply target for foldable OLED panels will rise from 8 million units this year by 87.5% to 15 million units next year; the iPad OLED panel supply target will increase from 2 million units to 9 million units.

Merrill SEC plans major innovation exemption for tokenized securities; may allow bypassing traditional trading platforms for direct on-chain trading

BlockBeats news: On September 9, The Rollup founder Andy posted that market rumors say the U.S. Securities and Exchange Commission (SEC) is preparing to introduce what would be the largest tokenization (Tokenization) innovation exemption to date. It could allow tokenized securities to be traded only through registered transfer agents, without broker-dealer licenses and without complying with rules related to traditional trading venues or ATS. It is also rumored that this could cover U.S. retail investors and overseas investors.

Andy said that if the above is true, the potential impact would be significant. Tokenized funds could issue and trade directly in the form of on-chain tokens, with transfer agents maintaining legal ownership records on-chain, while underlying assets such as stocks and bonds held by the funds could also be further tokenized—thereby forming an on-chain trading system of “fund token + underlying asset token.”

Andy later added that a large fund has already received an “approval” from the SEC, though this has not yet been officially confirmed. He speculates that ARK, Fidelity, or BlackRock could be potential participants.

If the policy is ultimately implemented, U.S. asset managers may accelerate the issuance of native equity tokens to compete for all-day liquidity and on-chain distribution channels, rather than waiting for third parties to mirror traditional securities into tokenized forms.

He further links this potential policy shift to the Trump administration’s recent push to open up crypto market regulation, as well as the CFTC’s efforts to bring perpetual contracts into the U.S. market, and believes that the U.S. regulatory environment may be gradually opening policy gateways for on-chain finance.