We name them after the people who noticed them. Moore's Law — named for Gordon Moore, the Intel co-founder who observed in 1965 that the number of transistors on a chip doubled roughly every 18 months. Swanson's Law — named for Stanford professor Richard Swanson, who documented that solar panel costs fell 20% with every doubling of cumulative production. Wright's Law — named for Theodore Wright, who found in 1936 that manufacturing costs fell predictably as cumulative output grew.
These are different industries, different technologies, different centuries. And yet the curves look almost identical. Each describes relentless, compounding cost reduction. Each has proven durable over decades. Each has driven transformations that remade civilization.
We call them laws, which implies they are forces of nature — inevitable, physics-driven, independent of human choice. They are not. They have a mechanism. And the mechanism, in every single case, is the same: competition.
The Mechanism: What Competition Actually Does
When companies compete for customers, they face a simple and relentless pressure: deliver more value for less money, or lose the sale. That pressure — multiplied across thousands of engineers, hundreds of firms, dozens of countries, year after year — is not abstract. It produces specific behaviors.
Companies that must compete invest in R&D to find cheaper manufacturing processes. They redesign supply chains to eliminate waste. They attract and concentrate talent by paying for breakthrough results. They share knowledge inadvertently as employees move between firms and as ideas diffuse through published research and industry conferences. Each incremental improvement is documented, builds on the last, and compounds.
This is the learning curve. The more you make something, the better you get at making it. And the better you get, the more you can sell, which means you make more, which means you get better still. The curve is not inevitable. It requires the pressure to keep improving. It requires competitors who will take your customers if you stop. Remove the competition, and the pressure evaporates. The curve flattens — or disappears entirely.
The improvement curves we name after scientists and engineers are not laws of nature. They are what free markets and competition produce, documented after the fact by people who were sharp enough to notice the pattern. The observation is theirs. The engine is ours — and only ours when we are free to compete.
Five Curves, One Engine
Semiconductors. When Gordon Moore made his observation in 1965, Intel did not yet exist. The semiconductor industry was already competitive — Fairchild, Texas Instruments, and others were fighting for the same customers in the same nascent market. The competition drove investment, investment drove improvement, improvement drove the curve.
Consider the contrast: in the 1960s and early 1970s, IBM dominated computing through long-term customer contracts, proprietary architectures, and a bundled sales model that effectively locked out rivals. IBM's mainframe customers faced no meaningful competitive pressure to improve. IBM did not face meaningful competitive pressure to improve. The result was decades of incremental progress, not exponential progress. When competition finally arrived — first through the minicomputer makers, then the PC revolution, then the open standards movement — the pace of improvement in every dimension accelerated dramatically. Not because the physics changed. Because the competitive pressure changed.
Today, 50+ billion transistors fit on a single chip. In 1971, Intel's first commercial microprocessor held 2,300. The price per transistor has fallen to essentially zero. Every doubling along that curve was driven by Intel vs. AMD vs. TSMC vs. Samsung — each spending billions to outcompete the others. The law is Moore's. The engine is competition.
Solar. Swanson's observation — 20% cost reduction per doubling of cumulative capacity — has held for nearly five decades. The mechanism is 10,000+ solar manufacturers competing globally across China, Germany, the United States, South Korea, and dozens of other countries, all fighting for market share in the world's fastest-growing energy market.
Compare this to the government-sponsored solar programs of the 1970s and 1980s. Solyndra, the US Department of Energy's flagship solar investment, absorbed over $500 million in federal loan guarantees and collapsed in 2011. It was not competing; it was being subsidized. It produced no learning curve. The private manufacturers competing for real customers — with real revenue at stake — drove the 90% cost reduction over 13 years that Swanson's Law predicted. The observation is Swanson's. The engine is competition.
Space launch. There is no official name yet for the curve that describes what happened to the cost of reaching orbit after SpaceX entered the market. But the numbers tell the story. The Space Shuttle — a government monopoly on US launch capability — cost approximately $54,000 per kilogram to orbit. It operated for 30 years with no meaningful cost reduction, because there was no competitive pressure to reduce costs. It was the only option. NASA paid whatever it cost.
Then SpaceX competed. The Falcon 9 brought the cost to $2,720 per kilogram — a 95% reduction. Starship, currently in testing, targets approximately $100 per kilogram. No government program, no scientific breakthrough, no new physics enabled this reduction. The same rocket equation, the same orbital mechanics, the same atmosphere. Different competitive pressure. When launch was a government monopoly, no curve existed. When competition arrived, the curve began immediately. The engine is competition.
Pharmaceuticals. The pharmaceutical generic market is perhaps the most direct real-world test of competition's mechanism — because it runs the experiment with a switch. While a drug is under patent, the manufacturer holds a monopoly. There is no competitive pressure on price. Prices stay high or rise. The day the patent expires, generic manufacturers enter. Prices fall 80–90% within months.
The same molecule. The same formula. The same manufacturing process. Entirely different competitive environment. Entirely different price curve. The GLP-1 revolution — the metabolic drugs that are transforming obesity treatment — will follow this same arc. As patents expire and competition enters, the drugs that currently cost hundreds of dollars per month will become accessible at a fraction of the price. The mechanism is identical to every other curve. The engine is competition.
Telecom. For most of the twentieth century, AT&T operated as a regulated monopoly. It held exclusive rights to the long-distance telephone network. The result was a rotary phone technology that remained essentially unchanged for decades. Prices did not fall. Service did not meaningfully improve. The monopoly had no reason to improve and no pressure to try.
The 1984 breakup of AT&T created the Baby Bells and opened long-distance to competition. Prices collapsed. Mobile phone technology emerged and proliferated at extraordinary speed. The internet era followed. None of this required new physics. It required competitive pressure. The breakup did not create new technology — it created competition, which created the pressure, which produced the improvement curves that followed.
