Moore Mindset and the Logistic Planet

Growing up in the ’00s and ’10s taught me a totally irrational view of technology. I learned to view any piece of tech, large or small, as a snapshot of a trajectory under constant acceleration. Take a machine, and anything measurable - speed, power efficiency, even beauty - will all improve exponentially with each successive generation. This has been the default for as long as I’ve been aware of tech, but it’s hard to blame myself for thinking this way. Moore’s Law predicts that the density of transistors on a computer chip will double every two years. I’m in my early 20s, and it’s been right for my entire life, plus a few decades before.

I never gave this much thought until a few weeks ago, when I saw an Apache helicopter.

The Apache is a beautiful machine. A sloped brow thickens vertically into an oblong thorax, flanked by narrow horizontal fins. The thorax tapers to a lithe tail long again as its body capped by a triangular fin. In flight, viewed from behind, its rotors are nearly invisible, existing mostly as a mental footnote to explain the technical wonder in progress. Look at a hawk for a moment and its flight is immediately legible as such. The Apache’s insectine body moves through the air in a totally different way, pitching and yawing around its center impossibly fast, as if it isn’t bound by the same laws that lift the hawk and carry it smoothly across a current. And the kicker - its form has been more or less unchanged since 1975. Like the crocodile, unevolved since the meteor strike that wiped out the dinosaurs, it’s a perfect predator.

I had never thought that we could have created a machine so beautiful and functional back then, let alone one that was still competitive in today’s marketplace of ideas. How could anything survive half a century of yearly doubling?

Earth, being a finite place, rarely hosts systems that grow exponentially. On a long enough timescale, almost everything here that looks exponential turns out to be logistic. A logistic curve starts at zero, and rises with an acceleration that suggests exponential growth. Then, suddenly, its growth rate slows precipitously and the curve flattens under a horizontal asymptote. The logistic curve describes the dynamics of almost every living population we’ve observed long enough in a habitat with a finite carrying capacity. A computer chip isn’t a habitat, but we’re starting to see the level-off, and unless you buy into wishful forecasts about quantum computing, we have a finite number of doubling events left.

Lately, of course, listening to tech hope-timism has been a recipe for disappointment. How many revolutionary innovations have ended up being vaporware? Peter Thiel famously lamented that we “were promised flying cars, but we got 140 characters.” Thiel and others like him turn this truth into a blame game, pointing the finger at their political heels like new hiring practices and “green” public investment standards. But maybe we were wrong to look at the rate of tech development in the 1990s and expect flying cars at all.

At the right cut-off point, the first-order approximation of a logistic curve looks like the “hockey stick” chart that gets investors out of bed in the morning. Investors love technology companies; in fact, I hope I find someone who loves me as much as investors love Uber - someone who’ll stick with me through thick and thin, who’ll say I’m worth more than Annheuser-Busch or UPS whether I’m lighting $3 billion on fire for three quarters in a row or I’m just plain revenue neutral.

Overall, the tech sector trades at a one and a half times higher price-to-earnings ratio than the rest of the stock market. The only way I can think to explain this is that investors have bought into the same Moore’s Law mindset that I did, and they’re waiting for the line to go vertical. But even in reality, the hyper-growth some tech companies enjoy is barely connected to returns. Moore Mindset was a bit of a hoodwink - the things that make a business profitable are mostly the same as they were before apps. Many tech innovations have actually been social and legal innovations that made the human labor that powers them more obscured, more precarious and less dignified. The algorithms are just icing on the cake.

The best expression of the mismatch here has been the large language model hype cycle. Their scaling laws require ever-increasing amounts of computational resources and human labeling work to drive linear performance improvements. We’re so past the point of diminishing returns that tech companies are starting to build datacenters with on-site nuclear power plants to meet some of their astronomical demand. Perhaps, not unlike the Apache, information processing is close to maturity. The increasingly speculative venture-funded hype that’s dominated the tech atmosphere lately would suggest so. Advances in computer hardware and telecommunications from the 1950s to the 1990s created a fertile “business habitat”, and IT may be reaching the top of its logistic curve.

This is happening at a time when the real world could use a lot more attention. Making the world and economy more resilient to the effects of climate change isn’t actually that technically complicated. We know exactly how to build more insulated buildings and how to farm to keep the soil alive. But private capital is less interested in that stuff, because it’s boring and it isn’t going to double every year forever. Our brains are so fucked by Moore Mindset that the only climate solutions we can imagine are tech solutions like a car that looks like an iPhone, is impossible to repair, and loves to catch on fire. Investors are even moving out of renewable energy because the returns aren’t exponential enough.

Logistic growth explains a lot of the incoherence in our current moment. It’s probably because we’ve never been near global carrying capacity, and so we’ve never confronted our own asymptote. People love to dunk on Francis Fukuyama for saying that history was over in the 1990s, but he was kinda right. Moore Mindset relies on infinite doubling, and we’re at the top. We need to let it go if we want to build a future.