According to a new Morgan Stanley note cited by Sawyer Merritt on X, analyst Adam Jonas says SpaceX’s neocloud strategy shifts costs to customers to build compute used to train its own robots, positioning the firm as a vertically integrated, full‑stack AI player.

According to Sawyer Merritt on X, SpaceX expects $35–$40 billion in Q4 2026 capex to nearly double compute capacity to about 3 GW with a sub‑12‑month payback.

According to Sawyer Merritt on X, citing Elon Musk, Grok Bot is experiencing the fastest product growth he has seen, with per-user usage rising, an enterprise version set for wide deployment in weeks, and expected to consume most SI compute.

According to Anthropic via X and the Simons Foundation, Anthropic pledged $5 million in funding, compute, and Claude access to support math AI workshops, conferences, and lectures managed by the Simons Foundation.

According to AnthropicAI, new metrics track AI-led R&D, agent oversight quality, and compute allocation, offering verifiable transparency for frontier AI.

According to timnitGebru, TechCrunch reports an NYU mathematician alleges OpenAI used Codex data to beat him to a major math proof after $22.5M compute.

According to SawyerMerritt, SpaceX’s AI compute hosting will exceed $3.4B per month by December, with a new $1.11B monthly deal starting Dec 1, 2026.

According to @CNBC, Trump’s AI executive order faces a key deadline as agencies finalize rules that could reshape model testing, reporting, and compute access.

According to SawyerMerritt, SpaceX revealed AI1 specs with 150 kW compute, vendor-agnostic chips, laser links via Starlink, and Bastrop-built solar tech.

According to SawyerMerritt, Raymond James rates SpaceX STRONG BUY with $800 target and sees AI orbit convergence driving massive revenue by 2031.

According to @ylecun, UN webcast spotlights AI safety governance, compute access, and open research priorities, per UN Web TV and UN Tech Envoy.

According to krea_ai, Krea launches a Creator Program offering compute, early access, rewards, and direct support for AI creators.

According to SawyerMerritt, Anthropic pays SpaceX about $41M per day for AI compute, totaling $15B yearly, with capacity ramping in 2026 per SEC filing.

According to SawyerMerritt, Anthropic will pay SpaceX $1.25B monthly for AI compute through May 2029, with discounted ramp in May–June 2026.

According to @OpenAI, Guaranteed Capacity offers contracted access to OpenAI compute for reliable long term scaling, backed by infrastructure investments.

According to SawyerMerritt, Musk outlines OpenAI suit views, SpaceX and Neuralink goals, and argues AI compute will move to space, per Forbes interview.

According to God of Prompt, Anthropic outlines compute gaps, chip smuggling, and model theft risks in a 2028 AI security brief, signaling political stakes.

According to AnthropicAI, the US and allies lead frontier AI and must secure chips, talent, and safety standards to maintain advantage, per its new paper.

According to The Rundown AI on X, today’s top AI stories highlight concrete product launches and business shifts: Google’s Gemini now has a native Mac desktop app, expanding multimodal assistants directly to macOS users and boosting enterprise adoption for on-device workflows (as reported by The Rundown AI). According to The Rundown AI, Notion introduced built-in Claude agents that automate audit and knowledge-work tasks inside workspaces, signaling deeper AI-native workflows for documentation and compliance. As reported by The Rundown AI, Snap plans to cut 1,000 jobs citing AI-driven productivity gains, underscoring cost efficiencies from automation across consumer tech operations. According to The Rundown AI, Allbirds is pivoting away from sneakers toward AI compute, suggesting a strategic reallocation into data center infrastructure and model-training demand. The Rundown AI also noted four new AI tools and community workflows, pointing to continued ecosystem expansion for developers and operators.

According to Ethan Mollick on X, Meta employees are competing to become “Token Legends,” ranking themselves by AI compute consumed, echoing the classic incentive risk warned in On the Folly of Rewarding A, While Hoping for B (Mollick shared the original paper link). As reported by The Information, internal leaderboards tie token usage to perceived productivity and influence, creating a status game where higher compute may signal impact (The Information). According to The Information, this metric could unintentionally reward excessive model calls over outcomes, raising cost, throughput, and model availability risks in large-scale LLM deployments. For AI leaders, the business opportunity is to implement outcome-aligned metrics—such as experiments shipped, latency budgets met, and unit economics per successful inference—while using governance controls like per-team quotas, cost dashboards, rate limiting, and evaluation harnesses to prevent compute gaming, as highlighted by The Information’s description of token-based status and Mollick’s incentive-design framing.