The Pitch
In 41 CE, the Emperor Claudius put 30,000 men to work for eleven years digging a six-kilometre tunnel to drain a lake. It silted shut; the lake stayed. In 1433, Ming China deliberately dismantled the greatest fleet the world had ever seen. Between 1969 and 1972, the Soviet Union launched its N1 moon rocket four times — and watched it explode four times. Wasted Utility is the ledger humanity never kept: the story of our costliest unrealized ambitions across more than two millennia, and the question none of these projects' backers dared to ask — what could we have built instead?
A reader could be forgiven for expecting another catalogue of famous failures. This is not that book. It rests on a fact-checked evidence base of more than twenty cases, each one quantified in modern dollars and checked against primary sources that run from Sima Qian's Shiji to SEC enforcement filings. What it measures is opportunity cost — the hospitals and harbours and cures the same money, and the same capable people, could have built instead. Money is not even the real unit of account here; human capability is. The 700,000 conscripts at the Qin Emperor's tomb and the 30,000 engineers writing token contracts at crypto's peak belong in the same column of the same ledger. And the finding underneath the cases is an uncomfortable one: the waste was rarely a matter of bad luck. It was overconfidence of a structural kind, visible years before the collapse it caused. The same pattern is running now, as hundreds of billions a year flow into AI data centres well ahead of proven returns.
None of this is an argument against ambition. The people in these pages had nerve to spare; what they skipped was the small, hard discipline of asking whether the bet still made sense before everything rode on it.
Nearly every project in here flashed a warning before it fell, and the people with everything at stake chose not to look. Catching that moment — a bold bet just as it turns bad — is the skill they lacked, and the one this book is trying to teach.
The Ledger
The book's chapters are drawn from the strongest cases in a completed research program. Each entry below is verified against primary sources; figures are midpoint estimates in 2024 US dollars. And because this is a ledger, every debit carries a credit — a real achievement of comparable cost in money and human effort, the thing that could have been built instead.
Qin Shi Huang's Mausoleum China
700,000 workers, 38 years, a tomb for one man. The dynasty fell three years after it closed.
Credit —The same conscript labour later dug China's Grand Canal: still moving freight 1,400 years on.
The Fucine Lake Tunnel Rome
30,000 men dug for eleven years to drain a lake. It silted shut; the lake outlived Rome itself, finally drained in the 1870s.
Credit —Claudius's own aqueducts, built in the same decade by similar crews, watered Rome for centuries.
Beauvais Cathedral France
The tallest vault in Christendom fell in 1284; the crossing tower followed in 1573. The nave was never built.
Credit —Roughly a Brooklyn Bridge, in modern terms.
The Ming Treasure Voyages China
The greatest fleet on earth made seven voyages — then the court turned inward and left it to rot.
Credit —A permanent Indian Ocean trade network: the one Portugal built decades later with a fraction of the ships.
The Darien Scheme Scotland → Panama
A fifth of Scotland's circulating capital sailed for Panama. Eight in ten colonists died; the Union followed.
Credit —A third of the Bank of England's founding capital — sunk in a swamp instead.
The Mississippi Bubble France
France's creditors traded the national debt for shares of a swamp. At the peak, one company was worth more than France's entire annual output — as if a single stock today out-valued US GDP. The fortune was paper; the colonists shipped to die at Biloxi, and generations of French distrust of banks, were real.
British Railway Mania United Kingdom
Around seven percent of GDP poured into railway schemes; roughly a third of the authorized lines were never built.
Credit —Dozens of London sewer systems — the Bazalgette works that actually ended cholera.
The French Panama Canal France → Panama
A sea-level canal that could not be dug. A billion francs lost and 22,000 dead before France gave up.
Credit —Roughly four times what it later cost the Americans to finish the canal for real.
The Soviet N1 Moon Rocket USSR
Four launches. Four explosions. The program was cancelled and erased from the official record.
Credit —Two James Webb Space Telescopes, with change.
The Superconducting Super Collider Texas, USA
Fourteen miles of particle-collider tunnel bored under Texas, then abandoned mid-dig by Congress.
Credit —CERN built the LHC — and found the Higgs boson — for not much more than America spent on the abandoned hole.
NHS National Programme for IT United Kingdom
The largest civilian IT program ever attempted delivered roughly 15% of what it promised.
Credit —Four Human Genome Projects.
Theranos California, USA
$724 million raised for a machine that never existed, beneath a nine-billion-dollar valuation.
Credit —Roughly half of what it cost to eradicate smallpox from the face of the earth.
Medupi Power Station South Africa
Costs nearly tripled, a unit exploded, and the lender expects no return — while the grid still fails.
Credit —Ethiopia built its Grand Renaissance Dam — comparable capacity — for roughly a quarter of the price.
The Crypto Experiment Global
The real debit isn't the trillions in paper losses — it's the minds. At its peak, crypto absorbed over 30,000 of the world's most capable engineers, plus the cryptographers, physicists, and quants behind them, chasing cheap money instead of hard problems. When the exodus finally came, the talent didn't return to medicine or energy. It moved to the entry below.
Credit —A Manhattan Project's worth of technical talent, pointed at token contracts.
The AI Buildout Global
Hundreds of billions a year into data centres, ahead of proven returns. The book's closing frame — and the reader's test case.
Credit —At ~$725B a year, nearly three entire Apollo programs. Annually.
*Paper valuations are recorded but not added to the total — the ledger counts only resources actually committed. Credit lines are order-of-magnitude comparisons against documented costs of real achievements, in 2024 dollars. Full ranges, sources, and normalization methodology are documented in the proposal; figures shown are midpoints.
Why This Book?
The Opportunity-Cost Lens
Most histories of failure explain what went wrong. This one fills the column they leave blank: what was forgone — the real problems the same money and human effort could have solved instead. No other book on the shelf is working that angle.
Original Cases, Not the Usual Suspects
Most books in this space retell the same handful of Western stories. This one ranges much wider — a Roman tunnel, a Ming treasure fleet, a South African power station that nearly broke a national grid — cases few readers, and few editors, will have seen before.
Timely and Urgent
The race to build AI data centres is committing historic sums against uncertain returns, and it is not alone — climate megaprojects and a new space age are doing the same. This book supplies the two-thousand-year base rate that separates a bold bet from structural overconfidence.
People at the Core
The book stays with people, not abstractions: the dreamers and the schemers, certainly, but also the ordinary people left standing in the fallout — a conscript at the Qin tomb, a settler bound for Darien, the young scientist who refused to stay quiet at Theranos.
The Evidence
The book is grounded in a completed research program rather than a thesis hunting for evidence. Every case has been checked against multiple independent sources, and costs are normalized to modern dollars by era-appropriate methods.
Sources range from Tacitus and the Ming Shilu to parliamentary records, stock-exchange archives, SEC filings, and peer-reviewed scholarship — with the central finding that in the great majority of cases failure was foreseeable: warning signs surfaced one to five years before collapse, and were overridden.
Outline
The High Price of Hope
A vast Roman tunnel that took eleven years to build and never did drain its lake. It sets the book's measuring stick: the cost of a failure, and the value of whatever it crowded out.
Monuments to Power — When Prestige Outruns Purpose
The Qin Emperor's tomb, Beauvais Cathedral's collapsing race for heaven, the Ming fleet ordered to rot at anchor. Political signaling: why rulers build past the point of reason, and who pays.
Phantom Wealth — The Anatomy of Manias
Scotland's Darien Scheme, John Law's Mississippi Bubble, Britain's Railway Mania. Herd behaviour, credit, and the numbers behind the crashes — and why each generation insists this time is different.
Engineering Hubris — When the Plan Defeats the Physics
The French Panama Canal, the Soviet N1, the Super Collider's abandoned tunnel, the NHS's £12.7-billion IT program. Technological overreach and sunk-cost escalation: how organizations keep paying long after the evidence is in.
The Modern Mirage — Today's Risks, Tomorrow's Lessons
Theranos's impossible machine, Medupi's runaway megaproject, crypto's gravitational pull on a generation of engineering talent, and the trillion-dollar AI buildout that inherited it. Spotting structural overconfidence in real time, with a framework for smarter bets.
Rewriting the Future
The patterns, the warning signs, and the discipline of asking what else could this buy? before the money moves.
Comparable Titles
Wasted Utility draws on two recent successes while claiming ground that neither of them holds:
How Big Things Get Done — Bent Flyvbjerg & Dan Gardner (2023)
Flyvbjerg diagnoses why megaprojects overrun and how to plan better. Wasted Utility extends the question backward through recorded history and asks what the failures actually cost the societies that bore them — the counterfactual Flyvbjerg leaves on the table.
Boom and Bust: A Global History of Financial Bubbles — William Quinn & John D. Turner (2020)
Quinn and Turner map 300 years of financial manias. Wasted Utility widens the frame beyond markets to monuments, fleets, state programs, and technological crusades — showing the same machinery of overconfidence operating wherever resources concentrate.
In reach and readability it aims for the tradition of Harari and Diamond — serious evidence carried by human stories, with a provocative question running underneath. The general reader gets the sweep of it; the entrepreneur or the policymaker gets base rates in place of anecdotes.
About the Author
Scott Francis Winder
What could we build instead?
Every entry here is a lesson somebody already paid for. Together they teach one unglamorous skill — telling the bets worth staying in from the ones you fold before they pull the rest down.
The manuscript is in draft. You can read the introduction now; the full proposal, the annotated outline, and further sample chapters are available on request.