Paul Hennessy/SOPA Images/Sipa USA via AP Images
An aerial view of the 870-acre, 75-megawatt solar farm at Babcock Ranch, Florida, the first solar-powered town in the United States
Silicon Valley has long had an almost religious fixation on the concept of innovation. Venture capitalists and tech startup entrepreneurs are constantly portraying themselves as people breaking through the crust of establishment dogma to create the buccaneering new technologies of the future.
So it’s odd that the tech industry has been consumed by a series of irrational manias over the past decade or so. First was cryptocurrency, which was supposed to be a revolutionary new financial technology that would bypass Wall Street, but turned out to be a speculative orgy infested with fraud to its very marrow. Then there was the NFT craze, which was supposed to provide some new method of digital scarcity, but turned out to be a land rush–style speculation frenzy, except without the land.
And now we have “artificial intelligence,” which is supposed to revolutionize work by automating all kinds of tasks. Now, I have experimented a bit with AI, and of the three it certainly seems to have the most potential. But, so far, every demonstrated legitimate use case involves careful data supervision and expert oversight—a rather expensive labor-enhancing tool, not the super-capable virtual personal assistant that CEOs are slavering over, and would justify the trillions of dollars in market capitalization in AI companies.
Meanwhile, as film editor Michael Sweeney points out on Bluesky, a truly enormous economic sea change—on par with the Industrial Revolution itself—is unfolding within America’s (and the world’s) energy system. Fossil fuel power is on its way out, replaced by renewable energy so cheap that it’s catalyzing spectacular innovations in all manner of long-established industries. It doesn’t get a tenth of the attention of the AI bubble, but this is what real innovation looks like.
Just consider the investment in solar, wind, and green manufacturing. In 2023, the U.S. solar industry installed 32 gigawatts of capacity, a 50 percent increase from 2022. The wind industry hasn’t been quite so spectacular of late, but still increased from 118 gigawatts of capacity in 2020 to 141 gigawatts in 2022. Grid-scale battery installations roughly doubled in 2023 and are projected to double again in 2024, to over 30 gigawatts, as this White House fact sheet points out. And electric-vehicle sales have soared from 20,000 per month in 2020 to 90,000 in 2023 (or about a tenth of total vehicle sales that year).
Renewable energy is cheap and getting cheaper, especially solar. In places with large solar penetration, the power can be so abundant during the day that its price drops almost to zero. That’s a problem for a traditional grid, but also an opportunity. In addition to compensating with battery storage, scientists (in government as well as the private sector) and entrepreneurs have been cooking up all kinds of new business ideas harnessing this dirt-cheap power to revolutionize manufacturing markets long considered mature.
For instance: Here’s a company with a plan to replace fossil fuel industrial heat—which accounts for about a quarter of all energy use—with “hot rocks” heated with renewable power. Here’s a company building two green steel plants in Mississippi and Ohio, the first of their kind in the U.S., thanks to a federal grant. Here’s another company selling power line monitoring devices that can allow the grid to transmit up to 40 percent more electricity. And here’s a company proposing to use green energy to create iron out of the toxic “red mud” by-product of aluminum refining—while cleaning up the mess in the process. And those are just a few among thousands of examples.
Importantly, nearly all these developments were catalyzed or directly caused by government policy. The Inflation Reduction Act, CHIPS and Science Act, and bipartisan infrastructure law have directed a flood of subsidies at green energy and manufacturing, building on decades of federal support for green tech. Many of today’s breakthrough technologies reaching commercialization were originally developed in government labs, or with government grants. The Loans Programs Office in the Department of Energy is directing hundreds of billions of dollars in loans to the most promising green companies.
These stories should cast some doubt on the innovation narrative behind even the most successful tech industry businesses. Facebook remains useful as a sort of telephone directory, I guess, but at the cost of devouring half the journalism industry and driving innumerable baby boomers toward conspiracy theories and extremism. Google used to be a very useful search engine, but the company’s latest innovation, which involves incorporating AI capabilities, has been producing errors and irking users; by now, the company is either losing or has given up the war on SEO spam and distracting ads—and also devoured the other half of the journalism industry. Amazon’s product ratings are systematically faulty and its store is full of counterfeits and deceptive products, not to mention the company’s abusive labor practices and harmful effects on the environment.
As the Apperceptive newsletter points out, software and web companies are extremely unusual in that they generally have relatively little capital costs and almost no marginal cost. A building full of coders and some server racks is a lot cheaper than, say, an auto factory. And once you’ve developed Microsoft Word, for example, each additional copy costs almost nothing to create—but your profits remain stable, and are protected by a government-granted monopoly. In short, creating such a business is relatively simple—or was at least in the Wild West days of the internet when most of today’s tech giants were founded—largely due to first-mover advantage.
On the other hand, business innovation involving actual objects is hard, particularly in markets like energy, steel, and automobiles, where production processes have been developed and streamlined over centuries. Labor productivity in American steel, for instance, increased by about 500 percent just between 1980 and 2015. And we’re not just talking about improving on the cutting edge by a bit here and there, we’re talking about overhauling the entire production process all at once, with novel technology that improves on the status quo right from the jump.
These are the real hero innovators of our time: scientists nobody has heard of in labs ironing out the kinks in perovskite solar panels for a few more percentage points of efficiency; chemists and materials scientists working out new designs that greatly decreased the price of batteries; physicists and engineers working on heat pumps that can work in very cold temperatures; and on and on.
It might be more fun to watch a megalomaniac on stage in a turtleneck promising his app will save humanity. But what will actually power the economy of the 21st century is a bunch of largely unknown, unrecognized people doing thankless, difficult, and often boring technical tasks.
