🗞 OpenAI's Chip Independence, France's $112B AI Push, and Altman's Revenue Vision

🔎 The Latest on the AI Frontier:

• OpenAI Accelerates Custom Chip Development to Reduce Nvidia Reliance

• Altman Addresses AI Access Inequality, Projects $100B Revenue Target

• France Launches $112B AI Initiative Focusing on Healthcare and Training

• DeepMind CEO Reviews DeepSeek's Progress, Projects AGI Timeline

• Sutskever's SSI Startup Eyes $20B Valuation Without Revenue

• Meta Under Fire for Using Pirated Books in AI Training

• Other news you might find interesting

🔧 OpenAI advances plan to reduce Nvidia dependence with custom chip development

• Company to finalize first in-house AI chip design in coming months, targeting TSMC 3-nanometer fabrication with potential mass production by 2026.

• 40-person team led by ex-Google chip expert Richard Ho collaborates with Broadcom, focusing initially on inference rather than training.

• Project aims to strengthen negotiating position with suppliers amid massive industry investments ($60B by Meta, $80B by Microsoft for AI infrastructure in 2025).

🤖 OpenAI CEO Sam Altman shares concerns about AI's uneven societal impact and proposes novel solutions

• Altman acknowledges potential power imbalance between capital and labor, suggesting a "compute budget" concept to ensure wider AI access as costs for basic AI capabilities drop ~10x every 12 months.

• OpenAI targets $100B revenue by 2029 and seeks $40B in funding, while planning significant investments with partners - up to $500B for data infrastructure.

• Company signals shift toward individual empowerment and increased open-sourcing, while maintaining long-term Microsoft partnership despite previous AGI-related contract terms.

💰 France commits $112B to AI development with focus on healthcare and workforce training

• French President Macron announced €109B ($112B) AI investment plan, positioning France to compete with US's $500B "Stargate" initiative in the global AI race.

• Investment includes training program to prepare 100,000 young professionals in AI-related fields, more than doubling current capacity.

• Paris will host 3rd global AI summit (Feb 10-11) with 1,000+ participants from 80+ countries, including leaders from Microsoft, OpenAI, Google, Anthropic, and Sakana AI.

🔬 Google DeepMind CEO assesses Chinese AI firm DeepSeek's capabilities, predicting AGI within 5 years

• DeepSeek's model represents "best work" from China per Hassabis, but uses existing techniques rather than new scientific advances.

• DeepMind claims its Gemini 2.0 Flash models surpass DeepSeek's efficiency, while experts question DeepSeek's claims about low-cost development.

• Hassabis projects AGI development within 5 years, aligning with OpenAI CEO Altman's views, while noting need for societal preparation and risk mitigation.

💰 Ex-OpenAI scientist's startup SSI targets $20B valuation despite no revenue

• Led by Ilya Sutskever, Safe Superintelligence seeks 4x valuation increase from $5B last September, backed by Sequoia, a16z, and DST Global.

• Company prioritizes "scaling in peace" over commercialization, contrasting with OpenAI's $4B revenue.

• Valuation talks proceed despite DeepSeek's market impact, as SSI focuses on developing safe superintelligent AI systems.

📚 Meta faces lawsuit over using 82TB of pirated books for AI training

• Court records reveal Meta employees downloaded pirated content from shadow libraries for LLaMA training, despite internal ethical concerns.

• Company took steps to hide downloading activity from corporate infrastructure after Zuckerberg pushed to "unblock" the process in January 2023.

• Case joins growing trend of copyright lawsuits against AI companies, including actions against OpenAI and Nvidia for unauthorized use of books and videos.

🖨 More news you might find interesting:

• OpenAI expands European presence with first German office in Munich.

• Chinese AI app DeepSeek faces scrutiny from South Korean intelligence over data privacy and bias concerns.

• Mystery AI models 'Chocolate' and 'Kiwi' appear on LM Arena, showing advanced capabilities in SVG generation

• Windsurf's Wave 2 update integrates advanced AI models with practical developer tools.

• Federal judge mandates immediate deletion of Treasury Department data accessed by Department of Government Efficiency (DOGE) operatives, restricts future access to vetted civil servants.

• Bill Gates shares cautionary vision of AI's societal impact in Fallon interview.

• T-Mobile launches Starlink satellite connectivity with broad carrier compatibility.

• JD Vance debuts on global stage at Paris AI summit, advocating business-friendly AI approach.

• Elon Musk’s DOGE develops custom AI chatbot 'GSAi' for federal efficiency drive.

Have any feedback? Send us an email