Humanity, Energy & AI, Part II: Humanity's Energy Arbitrage with Information and Mind

Once messages stopped needing a body to carry them, distance stopped being geography and became a price. The Chappe semaphore outran a horse courier fifty times over; the 1866 transatlantic cable collapsed a ten-day ocean crossing into minutes; radio reached 83% of American households within two decades; and the internet drove the marginal cost of sending information toward zero, reaching two of every three people on Earth. Claude Shannon's 1948 paper then changed the question: once information itself could be measured, a machine could do more than carry a message. Part II traces that turn through Deep Blue's brute-force chess, Watson's open-ended recall, AlexNet's self-taught vision, the Transformer's leap in scale, and ChatGPT's adoption curve, roughly 100 million users in about two months, faster than any consumer application before it. We believe the pattern is the same trade this series began with, now run on the most valuable material humanity has to spend: its own judgment.

Table of Contents

SECTION I: ENERGY ARBITRAGE FOR INFORMATION DELIVERY

1. Signals: Fire, Smoke, and the Semaphore
2. Wires: The Telegraph, the Telephone, and the Atlantic Cable
3. Air: Radio and the Satellite
4. Bytes: The Internet and Near-Zero Marginal Cost
   ‍

SECTION II: ENERGY ARBITRAGE FOR THE MIND

1. To Compute: Deep Blue and Brute Force
2. To Think: Watson and the Open Question
3. To Learn: AlexNet and the Neural Network
4. To Understand: The Transformer and Scale
5. To Reason: ChatGPT and the Fastest Adoption on Record