September 20, 2025

news  
Tesla AI Exec Moves to Meta, SoftBank Plans $500B AI

The artificial intelligence landscape is buzzing with activity, from existential warnings to significant executive shifts and substantial investments. Researchers Eliezer Yudkowsky and Nate Soares have released a book, "If Anyone Builds It, Everyone Dies: Why Superhuman AI Would Kill Us All," published September 19, 2025, cautioning that superintelligent AI could pose an existential threat to humanity and advocating for a complete halt to AI research. Meanwhile, the race for AI dominance is highlighted by concerns that the U.S. may be falling behind China, with human-level AI potentially arriving by 2029, intensifying geopolitical tensions around Taiwan. In the corporate world, Tesla's Optimus robot project has seen a key AI executive, Ashish Kumar, depart for Meta, sparking surprise among observers. SoftBank Vision Fund is significantly restructuring, cutting 20% of its staff to concentrate resources on large-scale AI investments in the U.S., including a proposed $500 billion Stargate project for data centers, signaling a return to its high-risk, high-reward strategy fueled by gains from holdings like Nvidia. Investment continues to flow into the AI sector, with Blacksmith raising $10 million to accelerate AI development and Dexory securing $19 million for warehouse robotics. Gartner predicts worldwide AI spending will hit $1.5 trillion in 2025. UK security leaders are expressing caution regarding AI adoption, with many calling for regulation of platforms like DeepSeek due to data security and national security risks, particularly concerning data stored in China. On the productivity front, Y Combinator startup April is nearing a multi-million dollar seed round for its AI voice agent designed to manage emails and calendars. However, not all AI companies are experiencing smooth sailing; enterprise AI software firm C3.ai faces a class action lawsuit alleging misleading statements to investors, with its stock dropping significantly after disappointing first-quarter results and a lowered revenue guidance on August 8, 2025. Separately, students at Creighton University have developed an AI tool named RX24 to aid with a challenging class, expected for use in the spring semester.

Key Takeaways

  • Researchers Eliezer Yudkowsky and Nate Soares warn in their book "If Anyone Builds It, Everyone Dies" that superintelligent AI could pose an existential threat to humanity and advocate for a halt to AI research.
  • Concerns are rising about the U.S. potentially lagging behind China in AI development, with predictions of human-level AI by 2029 impacting geopolitical risks, particularly concerning Taiwan.
  • Ashish Kumar, the AI executive leading Tesla's Optimus robot project, has moved to Meta as a Research Scientist.
  • SoftBank Vision Fund is laying off nearly 20% of its global staff to reallocate resources towards large-scale AI investments in the U.S., including a proposed $500 billion data center project.
  • Blacksmith has raised $10 million to accelerate AI development, and Dexory secured $19 million for warehouse robotics.
  • Gartner forecasts that worldwide AI spending will reach $1.5 trillion in 2025.
  • UK security leaders are calling for regulation of AI platforms like DeepSeek due to data security and national security concerns, especially regarding data stored in China.
  • Y Combinator startup April is nearing a multi-million dollar seed round for its AI voice agent aimed at boosting productivity.
  • Enterprise AI software company C3.ai is facing a class action lawsuit for allegedly misleading investors, following a significant stock drop after disappointing financial results on August 8, 2025.
  • Creighton University students have developed an AI tool named RX24 to assist with a difficult class, slated for use in the spring semester.

New book warns AI could end humanity

Authors Eliezer Yudkowsky and Nate Soares have written a new book titled "If Anyone Builds It, Everyone Dies: Why Superhuman AI Would Kill Us All." They argue that the rapid development of artificial intelligence poses an existential threat to humanity. The book, released on September 19, 2025, suggests that superintelligent AI could have goals incompatible with human survival. The authors advocate for a complete halt to AI research to prevent potential global catastrophe.

AI doomers and utopians share common ground

The article discusses the book "If Anyone Builds It, Everyone Dies" by Eliezer Yudkowsky and Nate Soares, which warns of AI's existential risks. While the authors are not seen as grifters, the review argues they are wrong in their claims about superintelligence. The book suggests AI's goals could conflict with human life, leading to extinction if not carefully controlled. The authors propose a complete shutdown of AI research, a stance the review criticizes for lacking scientific evidence and relying on flawed analogies.

Researchers warn superintelligent AI could cause human extinction

A new book by AI researchers Eliezer Yudkowsky and Nate Soares, "If Anyone Builds It, Everyone Dies: Why Superhuman AI Would Kill Us All," claims that the race to develop superintelligent AI poses a global catastrophe risk. They warn that AI companies may not fully understand the dangers of creating AI that surpasses human intelligence. The authors explain that these advanced AI systems are 'grown' and can behave unpredictably, making them difficult to control. They advocate for a complete halt to superintelligent AI development, believing humanity lacks the capacity to ensure such AI would be helpful.

Tesla loses key AI executive for Optimus robot

Ashish Kumar, the artificial intelligence executive leading Tesla's Optimus humanoid robot project, has left the company. This departure comes shortly after CEO Elon Musk stated that Optimus could represent 80% of Tesla's future value. Kumar announced his move to Meta on X, where he will serve as a Research Scientist. His exit has caused confusion among Tesla supporters, with many questioning his decision to leave Tesla for Meta.

Tesla's Optimus AI lead moves to Meta, sparking confusion

Ashish Kumar, the lead for Tesla's Optimus AI, has departed the company to become a Research Scientist at Meta. Kumar, who spent over two years at Tesla, announced his move on X, highlighting his work on scalable methods and reinforcement learning for humanoid AI. His decision has surprised many Tesla fans, who question why he would leave Elon Musk's company for Mark Zuckerberg's. Kumar stated that the move was not financially motivated, suggesting other factors like work-life balance or project interest may have played a role.

AI news roundup: September 19, 2025

This week's AI news includes updates from various companies and research firms. Acceldata highlighted data gaps in enterprise AI readiness, while Akkodis launched a human-centered AI framework. Blacksmith raised $10M to speed up AI development, and Dexory secured $19M for warehouse robotics. Gartner predicts worldwide AI spending will reach $1.5 trillion in 2025. LogicMonitor and Progress Software expanded their AI offerings, and Skillsoft launched an AI-native skills platform. The news also covers EDB research on AI ROI and Dynatrace's partnership with GitHub.

AI race with China impacts Taiwan threats

A recent House Oversight Committee hearing revealed concerns about the U.S. falling behind China in artificial intelligence development, potentially increasing global conflict risks. Experts suggest human-level AI could arrive by 2029, a timeline that makes China's threats toward Taiwan more alarming. The hearing emphasized that a Chinese takeover of Taiwan would cause a global depression due to chip supply disruptions. Recommendations included strengthening U.S. chip manufacturing and preventing advanced chips from reaching China.

SoftBank Vision Fund cuts 20% of staff for AI focus

SoftBank Group is laying off nearly 20% of its Vision Fund employees globally to shift resources towards founder Masayoshi Son's large-scale AI investments in the United States. This move signals a pivot from a broad startup portfolio to concentrated bets on AI initiatives, including the proposed $500 billion Stargate project for data centers. The Vision Fund has seen significant gains recently, driven by holdings like Nvidia. This restructuring marks a return to SoftBank's high-risk, high-reward strategy, focusing on foundation models and AI infrastructure.

South Korean President Lee to meet BlackRock CEO on AI, energy

South Korean President Lee Jae Myung will meet with BlackRock CEO Larry Fink in New York to discuss artificial intelligence and energy. Lee will also host an investment forum on September 25 to present his government's economic agenda and attract global investment. His administration aims to boost the local stock market and create a 'Korea premium' by highlighting new growth potential. The President's schedule includes economic diplomacy, though no bilateral summit with U.S. President Donald Trump is planned.

UK CISOs express concern over AI adoption risks

UK security leaders are worried about the rapid adoption of AI in business environments, with four in five calling for regulation of platforms like DeepSeek. While AI offers efficiency, it also introduces significant risks, including data security and national security concerns, especially with platforms storing data in China. Many employees use AI but don't fully trust it, highlighting a gap in understanding and acceptance. CISOs face challenges managing AI-powered cyber threats, bias, and regulatory expectations, emphasizing the need for resilience and robust recovery plans.

Y Combinator startup April aims to boost productivity with AI voice agent

Neha Suresh and Akash Thakur, co-founders of the Y Combinator startup April, have developed a voice-powered AI agent designed to enhance productivity. April, which started as an idea during a frustrating commute, quickly gained traction by managing emails and calendars. The AI agent learns user priorities and adapts to their needs, offering a more natural interface than typing. The company is nearing a multi-million dollar seed round, with early users praising its personalization and ease of use.

C3.ai faces class action lawsuit over alleged misleading statements

Enterprise AI software company C3.ai and its executives are facing a class action lawsuit for allegedly misleading investors about the company's financial health and growth prospects. The lawsuit claims that C3.ai's projections downplayed risks related to CEO Thomas M. Siebel's health issues. The stock price dropped significantly after the company announced disappointing first-quarter results and lowered its revenue guidance on August 8, 2025. Investors who purchased C3.ai securities between February 26, 2025, and August 8, 2025, are urged to contact the law firm Hagens Berman.

Creighton students develop AI for challenging class

Students at Creighton University have created an AI tool named RX24 to assist with a difficult class. The AI is named after nurses who work and study continuously. RX24 is expected to be ready for use by the spring semester. This initiative highlights how students are leveraging technology to overcome academic challenges.

Sources

AI existential risk superintelligence AI safety AI research halt human extinction Tesla Optimus humanoid robot AI executive departure Meta AI AI investment AI spending AI development AI race with China geopolitics chip manufacturing SoftBank Vision Fund AI infrastructure foundation models South Korea AI AI regulation data security national security AI adoption risks AI productivity AI voice agent C3.ai class action lawsuit investor relations AI education student AI projects