In 2023, a senior executive responsible for decarbonizing a major transportation company asked me, with some exasperation, what drove the “madness on hydrogen.” Hydrogen proposals kept arriving for applications where the energy balance and economics looked dreadful. That conversation became What Drives This Madness On Hydrogen?, an attempt to explain why enthusiasm persisted despite increasingly obvious problems. Three years later, I would change the weighting of my answer. Confirmation bias, familiarity and loss aversion were real, but I put them too close to the beginning of the causal chain. The more important question is why companies, governments and institutions acquired such strong reasons to keep hydrogen politically viable that adverse evidence repeatedly failed to stop the pathway. The harder question is what kept hydrogen alive after the technical case had weakened. The full TFIE Strategy Briefing retrospective follows the fossil-fuel and automotive incentives, the institutional evidence and the places where stopping rules disappeared. The constraints themselves were not obscure. A 2003 thermodynamic critique of the hydrogen economy traced the electricity required to make, compress, transport, store and use hydrogen and showed why it was intrinsically much more energy intensive than using electricity directly. That basic disadvantage never went away. Green hydrogen could be made with renewable electricity, but doing so did not make it cheap enough to compete broadly with direct electrification. The economics reinforced the physics. Electrolysis wants extremely cheap electricity, but an expensive industrial plant also wants high utilization. The cheapest wind and solar hours are intermittent. Run only during those hours and the plant sits idle much of the year; raise utilization and electricity becomes more expensive through overbuilding, transmission, storage or purchases in higher-priced periods. I described that tradeoff in 2022 as the economic seesaw of green hydrogen. Meanwhile the electrolyzer stack was routinely confused with the whole plant, which also needs transformers, rectifiers, water treatment, cooling, purification, compression, piping, controls, safety systems and substantial grid infrastructure. By 2025, observed project costs were badly outrunning major institutional forecasts. If those constraints were already visible, continuing enthusiasm requires an explanation other than technical ignorance. The answer becomes clearer when hydrogen is viewed not just as an energy carrier, but as a way of preserving the value of existing assets, skills and institutions. Gas companies own reserves, pipelines, processing infrastructure and subsurface expertise whose strategic value declines sharply in a world where heating, transportation and much of industry electrify directly. Hydrogen offered an alternative future in which gas could become blue hydrogen, pipelines could remain molecule pipelines, carbon capture could gain another market and existing engineering competencies could stay valuable deep into the 2030s and 2040s. The people making those decisions were not incapable of understanding hydrogen economics. Much of the relevant engineering sat directly inside their organizations. Legacy automakers faced a similar problem. Battery-electric vehicles shift competitive advantage away from combustion chambers, injection systems, transmissions, emissions controls and engine factories toward cells, motors, power electronics, software and high-voltage architecture. Hydrogen combustion and e-fuels offered a future in which more of the incumbent industrial system remained useful. Protecting workers, suppliers and industrial regions is a legitimate policy concern. Preserving inefficient propulsion technologies to avoid confronting that transition is something different. Governments then created another layer of commitment. Once hydrogen targets, subsidy programs, manufacturing plans and infrastructure strategies existed, entire ecosystems organized around them. The European Court of Auditors’ review of EU hydrogen policy found that major targets had been set without sufficiently robust analysis, yet those targets had already shaped billions of euros of investment decisions and years of industrial planning. That is where the absence of a stopping rule becomes important. Expensive hydrogen justified production subsidies. Missing customers justified demand support. Underused filling stations justified more stations and vehicles. Pipelines without committed throughput were justified as infrastructure needed to stimulate the production and demand that would eventually fill them. Each intervention could be defended on its own while the original question — whether hydrogen should serve the application at all — quietly disappeared. Hydrogen still has important uses. Existing ammonia, methanol and chemical production must be decarbonized, and some iron-production pathways may create additional demand. Those are applications where the molecule itself has value and can justify paying the unavoidable premium for low-carbon hydrogen. They never required hydrogen to become a general-purpose substitute for electricity in heating, road transportation or routine energy storage. That is what has changed most in my view of the hydrogen boom. Cognitive biases helped people defend positions they had already taken, but the commitments came first: reserves, pipelines, factories, supplier networks, public targets, budgets and careers. Much of the hydrogen detour was not inexplicable madness at all. It was rational organizational self-interest producing irrational energy policy, compounded by institutions that repeatedly failed to force the original proposition back onto the table when the evidence turned against it. For the deeper evidence on how fossil-fuel strategy, automotive incumbency, public targets and institutional self-preservation kept the pathway alive after the economics weakened, read the full analysis in TFIE Strategy Briefing.