The Sequence Radar-Issue #923: Last Week in AI: AI’s Industrial Turn
TheSequence Jesus Rodriguez ● Covered by 2 sources
AI’s big story last week wasn’t a new model. It was chips, power, and who controls the pipes. NVIDIA, Anthropic, and a16z all moved on the industrial stack, not just the software.
Based on reporting by TheSequence, Jesus Rodriguez — read the original for the full story.
Summary, retelling and take written by AI under human oversight; images are AI-generated illustrations. How we work · Report an error
For years, AI coverage has been a scoreboard of model upgrades. Better reasoning. Higher benchmark scores. Another scaling trick. Last week shifted the frame. The most important moves were about who owns the infrastructure that turns AI from clever software into something more like an industry.
The sharpest signal was the reported $12.9 billion deal for NVIDIA to acquire Hugging Face. That is not just a buyout of a popular AI platform. Hugging Face sits at the center of the open-model world, where developers find models, swap datasets, publish evaluations, and build applications. NVIDIA already controls a huge chunk of the compute layer. Add Hugging Face and it would also sit closer to distribution — the place where developers discover what to use next.
That is a lot of power in one pair of hands. It could make the ecosystem move faster. It could also make it feel less neutral, which matters because Hugging Face’s appeal has long been tied to the sense that it was open infrastructure rather than a vendor’s toll booth.
Anthropic’s reported six-year, roughly $45 billion deal with NScale points to the same shift from a different direction. The headline number is huge, but the 460 megawatts is the part that tells the story. Frontier AI companies are no longer just renting cloud capacity. They are locking up power-plant-scale infrastructure years ahead, and that pushes them into the business of electricity, cooling, networking, real estate, debt, and depreciation as much as software.
Then came a16z’s new $1.1 billion Machine Age fund, aimed at chips, memory, networking, data centers, robotics, and energy. That is the cleanest possible sign that the old “software eats the world” era now needs a bigger stomach. Software still matters. It just now depends on steel, copper, concrete, and a lot of power.
The throughline is hard to miss. NVIDIA is reaching for distribution. Anthropic is locking down industrial compute. a16z is financing the physical base underneath both. The model still matters, but it is becoming one part of a much larger machine.
My take — AI-written commentary, not fact-checked reporting
This is what AI looks like when the hype stops being cute and starts needing transformers, land, and a power bill. The market loves to talk about intelligence as if it floats above the grubby bits; it doesn’t. Open models will keep winning hearts, but the real fight is now over who owns the rails, and that is usually where the fees show up.
Read more about this at: TheSequence