About this video
What the video explains. The binding constraint of the energy transition has flipped. For twenty years the problem was making clean energy cheap, and that problem is largely solved: battery cells leave Chinese factories near forty dollars a kilowatt-hour and solar modules have never cost less. Yet electricity bills in three American states are up more than twenty percent in a year, new reactors are late almost everywhere, and the companies building small reactors have lost most of their market value. The new gates sit downstream of the generator. The system, in five stages. Primary energy, where the 2026 constraint is not the fuel but the fuel cycle - who mines, converts and enriches it. Generation, the cheapest it has ever been, where the constraint is financing the first plant of any new design. Firming and storage, the boom segment, though multi-day storage is one company deep. Grid delivery, the bottleneck of the decade. And consumption, where AI data centres arrived years ahead of new supply and turned absorption and affordability into the political front line. Stages one to three got cheap; stages four and five now set the price. Fission. China approved eight more reactors in a single decision on July 31. The United States started construction on zero new large reactors in the period; its story is restarts, not starts. The loan-backed pipeline is real money - $17.5B offered for up to ten large reactors, and in October two further conditional loans for uprates at plants already running, $4.2B to Vistra and up to $5B to Constellation, taking federal nuclear support past $34B. But notice what the money buys: more power from reactors that already exist. There is still no binding utility order for a new large reactor. Fusion. Its richest year ever, and its physics unchanged. The video sets Commonwealth Fusion's case for ARC against Sabine Hossenfelder's objection, then grades both against five gates: tritium breeding, magnets, materials, power handling and the wall-plug balance. The delivery wall, including as a computing problem. The queues, the transformer lead times, the voltage engineering - and one item from the laboratory aimed at delivery rather than supply. In September the municipal utility in Chattanooga switched on a commercial quantum computer inside its own operation. At the end of September it and its vendor published three papers, and the papers are more modest than the press release: no quantum speed-up is claimed, the sampling comparison is one classical solver against itself with and without guidance, the test grids are textbook networks rather than Chattanooga's own feeders, and nothing runs on a live circuit. What is genuinely new is that a utility now treats the delivery wall as a computing problem. How it was made. The presenter is a computer-generated character with an AI voice, reading a script drafted with Comuvia's AI system and reviewed by an editor. This is not a real person, and the caption "Computer-generated presenter - AI voice" is on screen beneath the figure throughout. The video carries YouTube's altered-or-synthetic-content label. Every segment marked "original audio" is an unaltered recording from the channel named on screen. The cards, the analysis and the complexity judgement are Comuvia's editorial work. How to inspect it. Every number is named and dated on screen, with its source stated at the moment the claim is made, so each can be checked against the original rather than against our summary. Where two sources disagree, both are shown and the disagreement is left standing. Where the conclusion is ours rather than the record's, the video says so, and the complexity judgement is labelled as a judgement. This is research and analysis, not investment advice.