New technologies need political help to become the Model Ts of tomorrow
The writer is author of ‘Exponential’
In 1908, the first Model T Ford came off the production line in Highland Park, Michigan. Over the coming decades, they would revolutionise not just the automotive market but the shape of our economies, cities and working lives. The magic seemingly lay in how the cars became increasingly affordable: a tool for the average citizen, not merely the very wealthy. Those first Model Ts set buyers back about $27,000 in today’s terms. By 1923, they were rolling off the production line for about a quarter of the price.
That incredible deflation for a then-breakthrough technology was not magic, nor unique. Rather, it was a consequence of “learning by doing”: as Ford’s engineers built more cars, they got smarter, defter and more efficient. In coming down this experience curve, the Model T became cheaper and, ultimately, ubiquitous. Some years later, Theodore Wright, an aeronautical engineer, was the first to describe this relationship quantitatively. He spotted that for every doubling of production, the cost of producing a new airframe dropped about 15 per cent. Experience is really a gift.
Today, as our economies contend with the maelstrom of geopolitics, trade, climate change and the rising costs of basic goods, being able to access advanced technologies is critical. Consider systems that turn polluting carbon dioxide into carbon-neutral jet fuels, high-capacity 3D printing, advanced sensors and computer chips for AI applications.
Many suffer the same problems of the first Model T: they are expensive. To make them cheaper, their engineers need the same thing that Ford’s did: experience. Economies of scale can help, but the real secret is Wright’s Law, getting better at building each and every one.
Wright’s Law applies to many complex technologies: solar panels, wind farms, electric vehicles, genome sequencers, high-performance silicon chips, nuclear fusion reactors and quantum computers. Ultimately these technologies end up being deflationary. A transistor, the basic physical unit of a silicon chip, cost the equivalent of $1,500 six decades ago, yet is now measured in billionths of a dollar. The key to speeding down the experience curve is to grow demand: the faster you can produce, the quicker you learn, and the more rapidly costs come down.
Economies should learn from Elon Musk. He started by making saloon cars for well-heeled drivers. In building about 50,000 of them by 2017, he figured out how to make cheaper mid-range cars, which are now the most popular electric vehicles in the US. The Tesla Model 3’s cost curve closely follows that of the Model T Ford a century earlier.
Frontier, a consortium of blue-chip firms, is trying to do the same with carbon removal systems. These technologies will suck excess CO₂ out of the atmosphere using giant fans or chemical filters and store it out of harm’s way. This may eventually be a useful tool to help us stick within our global carbon budget but prices need to come way down. Frontier, which counts Microsoft, McKinsey and Stripe among its members, sometimes spends in excess of $2,000 per ton of permanent carbon removal. The goal is to drive demand up, make learning faster and send prices to $100 per ton.
As technologies become cheaper and more prevalent, they bring huge economic rewards. For governments, this presents an opportunity. What could actively nurture them down the learning curve? Do we have to rely on long-term thinking entrepreneurs? Or can policy complement these efforts?
At the turn of the century, the German government introduced incentives to drive the uptake of solar power. The huge demand stimulus ended up making solar panels cheap for all of us — today it is the cheapest form of electricity in history.
Policies that ensure demand for promising technologies today so we can get them cheaply at scale tomorrow are crucial. Appropriate interventions can range from incentives for buyers to purchase mandates, to investment in infrastructure or experimentation. Joe Biden’s invocation of the Defense Production Act, which could include the US government becoming a buyer of first resort for needed technologies, may be such a trigger. Guaranteed markets help foster entirely new industries, which deliver their own economic benefits.
Of course, such approaches are not without risk. Politicians may bet on the wrong technologies or favour cronies. But diligent application of good practice — reliance on scientific expertise as well as the technological domains where investors are taking risks — could keep poor behaviour in check. Applied thoughtfully, it can bring the future forward.
Gloss