xAI Projects $13B Revenue by 2029; Nvidia, OpenAI Build UAE Data Hub
Key Takeaways
- xAI projects $13B revenue by 2029, seeks $5B debt, and plans $18B data center investment.
- Nvidia, OpenAI, and others are building Stargate UAE, a data hub with 100,000 Nvidia chips.
- Figure AI plans to manufacture 100K robots and seeks $1.5B funding at $39.5B valuation.
- Anthropic launched Claude Gov for US defense and appointed national security expert Richard Fontaine.
- EleutherAI releases 8TB Common Pile v0.1 dataset of licensed and open domain text.
Top Stories
xAI projects $13B revenue by 2029, seeks $5B debt
On June 6, 2025, Bloomberg reported that xAI projects annual earnings of over $13 billion by 2029 and $1 billion in gross revenue by the end of 2025, with a Q1 gross revenue of $52 million and a $341 million EBITDA loss. xAI is also seeking investors for a $5 billion debt sale via Morgan Stanley and a $300 million share sale at a $113 billion valuation, planning to invest $18 billion in data centers.
Nvidia, OpenAI, others build Stargate UAE data hub
On June 6, 2025, it was reported that Nvidia, OpenAI, Cisco, Oracle, and SoftBank are collaborating with G42 to build Stargate UAE, a data center hub with 100,000 advanced Nvidia chips, scheduled to go online in 2026.
Figure AI plans 100K robots, seeks $1.5B funding
On June 6, 2025, Figure AI CEO Brett Adcock discussed the company's relationship with BMW and confirmed plans to manufacture and deploy approximately 100,000 robot units within four years. Figure AI is reportedly seeking to raise $1.5 billion at a $39.5 billion valuation, a significant increase from its $2.6 billion valuation in February 2024.
Anthropic launches Claude Gov for US defense
On June 5 and 6, 2025, Anthropic launched Claude Gov, custom AI models for U.S. defense and intelligence agencies, and appointed national security expert Richard Fontaine to its long-term benefit trust. OpenAI, Meta, Google, and Cohere are also reportedly pursuing defense contracts.
EleutherAI releases 8TB Common Pile v0.1 dataset
On June 6, 2025, EleutherAI released the Common Pile v0.1, an 8 terabyte AI training dataset of licensed and open domain text, developed over two years with Poolside, Hugging Face, and academic institutions. The dataset was used to train Comma v0.1-1T and Comma v0.1-2T, which EleutherAI claims perform comparably to models trained on unlicensed data, demonstrating that high-quality models can be built using openly licensed content.
Business AI Impact
Anthropic cuts Windsurf's Claude 3.x model capacity
On June 7, 2025, it was reported that Anthropic cut off nearly all of Windsurf's first-party capacity to all Claude 3.x models. Anthropic co-founder Jared Kaplan stated they are prioritizing capacity for sustainable partnerships and that it would be odd to sell Claude to OpenAI, which has reportedly been in talks to acquire Windsurf for about $3 billion.
AMD acquires AI inference chip startup Untether AI
On June 6, 2025, AMD acqui-hired the team behind Untether AI, a startup that develops AI inference chips. Untether AI, founded in 2018, had raised over $150 million in venture capital.
Anthropic launches Claude blog, partners with Apple
On June 6, 2025, Anthropic launched a blog for its Claude models and partnered with Apple to power a new “vibe-coding” software platform.
TechCrunch AI Sessions: Funding, ethics, and strategies
On June 6, 2025, the TechCrunch Sessions: AI event featured discussions on various topics, including strategies for startups to grow with foundation models, ethical challenges facing AI, the increasing risks of investing in growth-stage AI startups, and the importance of human oversight in AI. Speakers from DeepMind, Twelve Labs, Amazon, ElevenLabs, Databricks, CapitalG, Khosla, Accell, Odyssey, Linear, NEA, Google Cloud, Tanka, OpenAI, and Oracle participated.
Lattice CEO emphasizes human-AI balance in companies
On June 6, 2025, Lattice CEO Sarah Franklin discussed the importance of balance in companies embracing AI, emphasizing the need to retain human employees while augmenting them with AI assistants. She highlighted that trust is the most important currency and that humans must have oversight of any AI technology implemented by a company.
Ethical AI Concerns
UK court warns lawyers on AI-generated fake cases
On June 6, 2025, London's High Court warned lawyers that using AI to cite non-existent cases could lead to contempt of court or criminal charges. A senior judge criticized lawyers for using AI tools that referenced fake case law and emphasized the ethical obligations of lawyers to not mislead the court.
Trending AI Tech
Google Gemini Pro/Ultra gets 'scheduled actions' feature
On June 6, 2025, Google launched 'scheduled actions' for Gemini AI Pro and Ultra subscribers, enabling the AI to perform tasks at specific times. Users can manage these planned tasks within the Gemini app settings.
ChatGPT introduces meeting recording and connectors
On June 6, 2025, it was reported that ChatGPT introduced meeting recording and connectors for Google Drive, Box, and more.
ChatGPT offers similar scheduled actions feature to subscribers
On June 6, 2025, it was noted that OpenAI's ChatGPT provides a comparable feature for subscribers, allowing the AI chatbot to send reminders or perform recurring actions.