Meta builds AI ‘M-Team’ in AI Talent War: Meta Founder/CEO Mark Zuckerberg is driving an accelerated effort to build it’s AI capabilities beyond Meta AI and open source Llama. Especially relative to industry peers Google, Microsoft/OpenAI, Elon Musk’s xAI/X/Grok/Tesla et al. And heightened competition from AI companies in China like DeepSeek, Alibaba et al. After its 49% investment in AI Data leader Scale.ai and bringing on its founder/CEO Alexandr Wang to head Meta’s ‘Superintelligence’ team the company is in talks to bring on Github founders and VCs Nat Friedman and Daniel Gross as part of its’ AI Management Team, aka ‘M-Team’. Meta also held talks to acquire Perplexity under similar terms before Scale.ai, and has discussed hosting Ilya Sutskever’s Safe Superintelligence in Meta AI data centers. It’s clear that Zuckerberg is thinking ‘out of the box’ in expanding Meta’s AI envelope. The company also announced upgraded AI ‘Smart Glasses’ with Oakley, expanding its portfolio beyond Ray-Ban Smart Glasses. Meta Founder/CEO Mark Zuckerberg is increasing his overallI ambitions across the board. More here and here.
OpenAI & Microsoft close to an impasse: Several reports this week on OpenAI and Microsoft increasingly at a negotiation impasse on its iconic partnership adjustments. Issues range from Microsoft’s blanket access to OpenAI IP through 2030, to its ownership stake in the for-profit entity being negotiated with existing stakeholders. Also on the table are revenue split arrangements on various lines of AI products and services by both OpenAI and MIcrosoft. These included API driven businesses to developers and businesses, as well as OpenAI’s fast growing ChatGPT subscription businesses for AI Reasoning and AI Agent products. Also on the table are OpenAI’s options to build its own AI Infrastructure beyond Microsoft Azure. Both CEOs Satya Nadella and Sam Altman also seem keen on continuing other elements of their partnership intact.. More here.
Apple’s viral paper on AI Reasoning : A new AI paper from Apple has sparked a debate over AI Reasoning innovations that are driving the efforts of major LLM AI companies. The paper titled ‘The Illusion of Thinking’ posits the reasoning capabilities anticipated with current AI techniques may be overestimated. This counters other essays by OpenAI’s Sam Altman that we may be close to ‘A Gentle Singularity’. Most AI Researchers agree that AI Scaling across several dimensions remains promising ahead, both for AI Reasoning and AI Agents, to AIs also capable of original innovations. But the progress is stop and go, with speed bumps along the way. . More here.
Elon Musk raising more capital for xAI/Grok: Elon Musk is reportedly raising more capital via debt and equity to fund AI Infrastructure spend for his xAI/Grok projects. At this point Elon is competing against an ‘AI Arms Race’ vs Meta’s Zuckerberg, as well as the well documented AI capex budgets of Google, Microsoft, OpenAi and others. Not to mention the Ai spend by companies in China. The quest also includes securing additional distribution and data sources for Grok beyond X/Twitter, as well as possible customers on the enterprise side of the docket. Elon’s recent deal with Microsoft Azure is an effort to expand in that direction. Not to mention of course the big AI infrastructure that Tesla also continues to need in its efforts for FSD (full self-driving cars), as well as AI Robotaxis and humanoid Optimus Robots. More here.
AI Fears persist despite current AI investment boom: Recent reports continue to underline the ongoing concerns over the downsides of AI, with statements by industry leaders at Anthropic and beyond. Known as ‘p(doom)’ scenarios, the industry is having to continue to gauge the probabilities of AI ‘going wrong’ vs going right. Problem here of course continues to be to handicap the ‘unknowable’ elements, especially when the underlying efficacy of AI technologies are still being fine-tuned and debated at the same time. Despite the hundreds of billions being poured into AI Infrastructure and AI Talent to scale the efforts at levels unprecedented via prior tech waves. More here.
Other AI Readings for weekend:
New MIT ‘SEAL’ methodology to LLM AIs to ‘keep learning on the fly’. More here.
Softbank’s Masayoshi San pitching Trump WH a China, Shenzhen style $1 trillion AI/Robotics hub dubbed Crystal Land in Arizona. More here.
Up next, the Sunday ‘The Bigger Picture’ tomorrow. Stay tuned.
(NOTE: The discussions here are for information purposes only, and not meant as investment advice at any time. Thanks for joining us here)