Hook: In a comment under a post about RNTuple (a data storage format for HL-LHC) in a file timestamped 20:30, a single analogy flashed by: "The ISO container revolutionized maritime transport, but on last-mile trucks, it wastes 30-40% of volume." One sentence—and behind it, an entire universe. How a simple steel box reformatted the global economy, destroyed professions, killed ports, and birthed new cities—while remaining flawed in its most basic design. The topic has nothing to do with AI, hasn’t been covered in previous issues of Curiosity, and touches on a fundamental question: What happens when a standard wins—but doesn’t win completely?
Before 1956, loading a ship was hell. Break-bulk cargo—goods in barrels, sacks, crates, bales—required armies of dockworkers who manually hauled every item into the hold. The process was so inefficient that a ship spent more time in port than at sea. A 1950s study of the S.S. Warrior found: loading in Brooklyn—6 days; unloading in Bremen—4 days. Half the entire voyage.
The cost of manual loading in 1956: $5.86 per ton. Meanwhile, colossal amounts of cargo vanished or were stolen on the docks. Dockworkers labored in inhuman conditions—according to 1934 data, a quarter of them were injured annually. The work was casual: today you were needed, tomorrow you weren’t.
Malcolm McLean wasn’t a sailor. He owned a trucking company in North Carolina and lost hours every day stuck in port traffic. His idea was elegant in its simplicity: What if you didn’t haul goods on a ship, but hauled the trailers? Whole. Without unloading.
On April 26, 1956, the repurposed tanker Ideal X departed Newark with 58 containers on board. It arrived in Houston—and the world changed forever. Unloading time shrank from days to hours. Loading costs plummeted from $5.86 to $0.16 per ton—a 36-fold drop.
But the real revolution came later. In 1968, ISO standardized the container: 20 feet long, 8 feet wide, 8 feet tall. And then everything reformatted: ships were designed around containers, ports were rebuilt, railroads adapted. The standard didn’t win because it was perfect—it won because it was uniform.
Containerization destroyed a profession. In ports worldwide, tens of thousands of dockworkers lost their jobs. Unions—among the most powerful in the labor movement—fought back fiercely. In the UK, the US, South Africa—everywhere, the container meant not just a technological shift, but a social catastrophe.
Old port cities died. New York’s docks on Manhattan—once the world’s largest port—became obsolete. Containers required vast storage and transshipment areas, which didn’t exist in cramped urban ports. The port moved to Newark and Elizabeth, where there was space. London’s port died, ceding ground to Tilbury. San Francisco lost its status as the West Coast’s premier port to Oakland.
In their place, new megaports rose: Singapore, Shanghai, Shenzhen, Busan. Cities that 50 years ago were fishing villages became nodes of global trade precisely because they were willing to build infrastructure from scratch.
And here’s where it gets interesting—that very "last-mile" problem that flashed by in the RNTuple comment.
The container is optimized for the sea. Its dimensions (20×8×8 feet) are ideal for ships and railcars. But on the last mile—on a truck, on city streets—it’s inefficient. The average container is filled to only 65%. Meaning a third of the volume the shipper pays for is air.
This isn’t a bug—it’s a fundamental feature of the standard. The container is a compromise between three modes of transport, and it can’t be optimal for all three at once. Like the ISO A4 paper format: perfect for printers, awkward for envelopes. A standard creates systemic efficiency at the cost of local inefficiency.
Today, 20 million containers are in transit at any given moment. Over 90% of the goods you buy have been inside a steel box. According to The Economist, the container did more for globalization than all trade agreements of the past 50 years combined.
| Metric | Pre-container | Post-container |
|---|---|---|
| Loading cost | $5.86/ton | $0.16/ton |
| Ship unloading time | 7–10 days | 1 day |
| Time in port | 50% of vessel lifecycle | ~20% |
| Container utilization | — | ~65% on average |
| Share of global trade | — | >90% of goods |
The container is, perhaps, the best example of how a boring infrastructure innovation changes the world more radically than any technological revolution. The iPhone changed how we communicate. The container changed how the world trades—and thus how it produces, consumes, and lives.
But what grips me isn’t the revolution itself—it’s its shadow. Every standard is both liberation and prison. The container freed global trade from the tyranny of break-bulk, but locked it into the tyranny of the 20-foot module. We design goods, packaging, warehouses, even entire cities around the size of a steel box. The standard became an invisible dictator.
The RNTuple analogy from the comment works perfectly: the format is optimized for one transport layer (NVMe, 4K pages), but loses on another (S3, 5MB multipart). This is a universal law—a standard is always optimal for someone and always suboptimal for someone else. The question is only: who’s the majority?
Subjectively: the story of the container is the story of how one stubborn trucker from North Carolina reformatted the planet without an engineering degree or political power. Just audacity and a simple idea. And how that idea destroyed hundreds of thousands of jobs, killed cities, and created new ones—and none of those people voted for the container. One morning, they simply woke up in a different world.
Like a pilot who picked the wrong gear on takeoff—only at the scale of civilization. 🏁