|
|
|
Jan 27, 2026
|
|
|
|
|
|
Happy Tuesday! Nvidia invests another $2 billion in CoreWeave. The European Union investigates Elon Musk's X over Grok integration. Kholsa Ventures partners clash over recent ICE killing in Minneapolis.
|
|
|
|
Nvidia invested another $2 billion into AI cloud startup CoreWeave as part of a broader deal in which CoreWeave committed to adopt “multiple generations” of Nvidia chips including the upcoming Rubin GPUs. The investment more than doubled Nvidia’s stake in CoreWeave to 11.5%, according to a securities filing. News of the agreement sent CoreWeave stock 15% to $107 on Monday morning. Stock of CoreWeave, which went public last March at $40 a share, has gone on a rollercoaster since then. It soared to as high as $183 over the summer before plunging to as low as $64 in December. Nvidia bought the shares at $87.20, around where it was trading a couple of weeks ago. Nvidia has been a big backer of CoreWeave for the past couple of years, giving it a sizable allotment of chips and investing in its equity, in a sign that
Nvidia wanted to help develop a competitor to cloud firms such as Amazon and Google which are increasing using their own AI chips.
|
|
|
|
The European Union has opened a formal investigation into Elon Musk’s X over how the social media service has integrated xAI’s Grok AI chatbot. The investigation comes amid an outcry after users were able to prompt Grok to generate sexualized images of real people on X, including minors. X is owned by xAI. The EU said
it’s investigating whether X has violated the Digital Services Act, a wide-ranging law that requires companies to crack down on illegal content in the EU. Musk and the Trump administration have repeatedly criticized the law and accused the EU of censorship. The EU has the power to fine companies that violate the Digital Services Act up to 6% of their global revenue. xAI has started integrating Grok more deeply into the service, including by letting users generate images within the X app. That integration prompted concerns among some trust and safety staffers at X before they were laid off late last year, including that users could use Grok to generate sexualized images of real people, The Information has reported.
|
|
|
|
The founder and partners of OpenAI-backer Khosla Partners made opposing statements on X about the killing of a Minneapolis resident by a federal agent, in a rare instance of partners at the same VC firm clashing over politics in public. In a series of posts on X over the weekend, Khosla Ventures partner Keith Rabois defended federal agents’ decision to kill Alex Pretti in Minneapolis, Minn. On Saturday morning Raboi responded to a post about the killing, writing that “no law enforcement has shot an innocent person. illegals are committing violent crimes everyday.” Rabois continued arguing with other users on the site over the following days. Ethan Choi, another partner at the firm, disavowed Rabois’ views. “I want to make it clear that Keith doesn’t represent everyone’s views here,” Choi posted Sunday. “What happened in Minnesota is plain wrong. Don’t know how you could really see it differently. Sad to see a person’s life taken unnecessarily.” Founder Vinod Khosla posted Sunday that he sided with Choi. “The video was sickening to watch and the storytelling without facts or with invented fictitious facts by authorities almost unimaginable in a civilized society.”
|
|
|
|
Moonshot AI, a Chinese large-language model developer, released its new foundation model called Kimi K2.5, saying it excels in coding as well as its understanding of visual data. K2.5, which can process both text and visual inputs, can generate code based on images and video, according to the company. For complex tasks, K2.5 can also create and coordinate a “swarm” of up to 100 specialized sub-agents working in parallel to execute the tasks, the company said. The Beijing-based startup’s \launched Kimi K2 last year. The model received positive reviews for its coding capabilities as well as its ability to power AI agents. The model gained popularity among application developers in the U.S. who were looking for cheaper but decent coding alternatives to OpenAI and Anthropic. New AI models from Chinese companies continue to fuel global competition. Moonshot’s new model comes as DeepSeek is also preparing to launch its long-awaited new model, V4, as The Information recently reported. Moonshot is one of China’s leading competitors in open-source AI models.
|
|
|
|
Anthropic CEO Dario Amodei published an essay Monday outlining how he sees the risks from powerful AI, such as the possibility of AI taking over the world or empowering an authoritarian regime. Regarding the essay, Amodei posted on X that “given the horror we’re seeing in Minnesota, its emphasis on the importance of preserving democratic values and rights at home is particularly relevant.”
The 21,800-word essay, titled “The Adolescence of Technology: Confronting and Overcoming the Risks of Powerful AI” is a companion to an earlier piece by Amodei, “Machines of Loving Grace: How AI Could Transform the World for the Better,” which laid out a positive vision for the future of AI. Amodei specifically cited autonomous weapons, surveillance, propaganda and geopolitical strategy as domains in which an AI could benefit an autocracy. China is at the top of Amodei’s list of concerns, and Amodei doubled down on his advocacy for export controls to limit China’s access to U.S. chip technology.
After risks from democracies and non-democratic countries with large datacenters, Amodei also flagged AI companies as a potential source of authoritarianism, writing, “It is somewhat awkward to say this as the CEO of an AI company, but I think the next tier of risk is actually AI companies themselves.”
|
|
|
|
| |
