Global Vista Insights
Subscribe
GVI World

Why Software Ate the World but Rockets Didn't

Computing power has compounded for fifty years. Airplanes still fly at 1970 speeds. That asymmetry needs explaining.

2 min read

The laptop I’m writing this on has more computing power than the Apollo Guidance Computer by a factor of roughly 100 million.1 That is not a metaphor — it is a measurable, documented fact about what fifty years of compounding in silicon has produced.

The applications have compounded across every domain of economic life. Smartphones, modern data centres, banking systems run entirely on software, trading and pricing of every asset class globally — all of it rides on the back of this progress.

But progress is not evenly distributed.

Commercial airplanes cruise at roughly the same speed they did in the 1970s.2 Infrastructure, particularly housing, is built in largely the same way — concrete, steel, manual labour — with incremental efficiency improvements at the margin. Schools operate on a model recognisably similar to what existed a century ago. Even gains in life expectancy, which were dramatic through the 20th century, have begun to plateau in much of the developed world.3

So why has progress been so heavily weighted toward bits?

One answer is the economics of replication. Software, once written, costs almost nothing to copy. The marginal cost of delivering one more unit — one more app download, one more search query, one more streamed film — approaches zero. That is not true of a bridge, a hospital, or a commercial aircraft. Physical infrastructure has high capital costs, long build cycles, heavy regulation, and no equivalent of the “deploy to a billion users overnight” moment that software takes for granted.

A second answer is that the atom-world has already absorbed a great deal of progress that is now invisible. The steel in a modern building is vastly superior to Victorian-era steel. Jet engines are dramatically more fuel-efficient than their 1970s equivalents — even if the plane flies at the same speed. Modern pharmaceuticals extend and improve life in ways that would have seemed miraculous fifty years ago. The gains are real; they are just harder to see because they show up as reliability, safety, and efficiency rather than raw speed.

The asymmetry matters because capital follows it. Venture capital, talent, and corporate R&D have concentrated in software and adjacent fields for two decades. The atom-world — energy, housing, transport infrastructure, manufacturing — has been relatively starved of the same compounding attention. Whether that rebalances, and how quickly, is one of the more consequential economic questions of the next generation.

  1. The Apollo Guidance Computer operated at approximately 0.043 MHz with 4KB of RAM. A modern laptop operates at ~4–5 GHz with 16–32GB RAM — a difference of roughly 100 million times in raw compute. NASA Technical Note D-4154 (1968); see also Charles Fishman, One Giant Leap (2019).

  2. Average cruising speed of commercial jets has remained ~900 km/h since the Boeing 747 entered service in 1970. FAA Aerospace Forecast data; see also The Economist, “Why flying isn’t getting any faster,” 2019.

  3. US life expectancy peaked at 78.9 years in 2014 and declined through 2021. CDC National Center for Health Statistics, National Vital Statistics Reports, 2022.