The progress in deep learning, from AlexNet's GPU leap to today's massive models, is best understood as a history of scaling compute. This scaling, resulting in a million-fold increase in power, enabled the transition from text to more data-intensive modalities like vision and spatial intelligence.

The progression from early neural networks to today's massive models is fundamentally driven by the exponential increase in available computational power, from the initial move to GPUs to today's million-fold increases in training capacity on a single model.

The International Energy Agency projects global data center electricity use will reach 945 TWH by 2030. This staggering figure is almost twice the current annual consumption of an industrialized nation like Germany, highlighting an unprecedented energy demand from a single tech sector and making energy the primary bottleneck for AI growth.

For years, the tech industry criticized Bitcoin's energy use. Now, the massive energy needs of AI training have forced Silicon Valley to prioritize energy abundance over purely "green" initiatives. Companies like Meta are building huge natural gas-powered data centers, a major ideological shift.

The U.S. has the same 1.2 terawatts of power capacity it had in 1985. This stagnation now poses a national security risk, as the country must double its capacity to support AI data centers and reshoring manufacturing. The Department of Energy views solving this as a "Manhattan Project 2.0" level imperative.

Instead of relying on hyped benchmarks, the truest measure of the AI industry's progress is the physical build-out of data centers. Tracking permits, power consumption, and satellite imagery reveals the concrete, multi-billion dollar bets being placed, offering a grounded view that challenges both extreme skeptics and believers.

Satya Nadella clarifies that the primary constraint on scaling AI compute is not the availability of GPUs, but the lack of power and physical data center infrastructure ("warm shelves") to install them. This highlights a critical, often overlooked dependency in the AI race: energy and real estate development speed.

Most of the world's energy capacity build-out over the next decade was planned using old models, completely omitting the exponential power demands of AI. This creates a looming, unpriced-in bottleneck for AI infrastructure development that will require significant new investment and planning.

AI's computational needs are not just from initial training. They compound exponentially due to post-training (reinforcement learning) and inference (multi-step reasoning), creating a much larger demand profile than previously understood and driving a billion-X increase in compute.