SpaceX is making minihard, an improved modular 220K GB300 data center. It will be small building to the right of Macrohard and Macrohardrr. It is the most important thing Elon and SpaceXAI has said about its infrastructure this year. Minihard is about 25% of the floor area of Macrohardrr but the same 220K GB300s. Minihard …
When you use an AI to write code, you pay for tokens — chunks of text the model reads and writes. Frontier models charge a lot per token. Cheaper models charge a fraction. Think of building a house. An architect charges $500 an hour; a bricklayer charges $40. Both are essential. But you’d go broke …
Cursor has run experiments to test the limits of scaling agents to cooperate toward a goal. Their hypothesis is they can unlock a new tier of task scale and complexity. They are looking to speed up by thousands of times from 1000 commits per hour to 1000 commits per second. Tools like Git and Cargo …
Grok 4.5 (1.5 trillion parameter) at 11 will likely crack the top ten LLM on open router. Based on the number of models they are releasing. They could have new models in the 50% frontier coding score but at prices lower than Anthropic for Opus 5. Later versions of Grok 5.X could push out Opus …
A highly detailed, month-by-month financial model for xAI/SpaceX’s AI data center leasing business. This highlights three existing major leasing deals already in place worth $27.8 billion annualized. Google deal ramping to nearly $1 billion per month (fully online by October) and predicts 1–2 additional major leasing deals will be announced before the end of the …
The CEO Supermicro @Supermicro, Charles Liang, said Supermicro will help build a new gigawatt size AI data center for SpaceX and it will be the fastest build they have ever done. The Time to Operation will be shorter than the 1 year for colossus 2. @elonmusk A big congratulations to SpaceX’s greatest IPO and proud …
SpaceX AI leasing deals worth $28 billion annualized and earn roughly 3–5x more revenue per MW than traditional multi-year cloud commitments. SemiAnalysis explains that SpaceX is selling IMMEDIACY. If you need a lot of AI data center then you can get it now or in months. Anthropic got 300 MW usable within a month, versus …
Gavin Baker says anything that increases competition and compresses margins at the model layer (foundation models) is net positive for every other layer in the AI stack — power, semiconductors, hyperscalers/neoclouds, and even software. A world with only 2–3 dominant high-margin frontier labs is bad for the rest of the ecosystem because those labs become …
The most misunderstood line item in the AI buildout is depreciation. Every hyperscaler is reporting compressed margins as tens of billions of dollars of AI capex hit the income statement through D&A while revenue is still ramping. Amazon’s own accounting illustrates it: roughly $220B of AWS-attributable 2027 capex, 70% in short-life assets (AI servers on …
SpaceXAI has real-world proof of 6–12 month deployments for large clusters via warehouse retrofits, phased build-out, and on-site/mobile power. They have gas turbines from partners like Solar Turbines + APR Energy that can install in days to 6-12 weeks. People are worried about how quickly the big tech hyperscalers are spending their hundreds of billions …
Looking at statistics from openrouter shows Grok 4.5 could capture a lot of AI spending market share by dominating the value AI segments. Brian WangBrian Wang is a Futurist Thought Leader and a popular Science blogger with 1 million readers per month. His blog Nextbigfuture.com is ranked #1 Science News Blog. It covers many disruptive …