In August 2011, Marc Andreessen wrote a Wall Street Journal essay called “Why Software Is Eating the World.” The thesis has aged into received wisdom: every industry would be reshaped by software, and the companies producing it would inherit the economy. He was right. In the last fifteen years software has consumed every problem big enough to justify the cost of producing it.
The price of building software
Up until a year ago, it was a hard, expensive, and difficult process to create software that solved even the simplest problems. Investors became obsessed with TAM because there just wasn’t enough money to be made in solving most problems. A year later, we see that idea collapsing because the ability to produce software is approaching zero. Every small business, every gym, every individual professional can now build bespoke software in an afternoon that they can use once or iterate on forever. That matters because most of the global economy is still offline.
It’s easy to forget that most processes in most industries are still analogue, but that paradigm is shifting. By the end of our lifetimes, I’d wager that changes and there is an enormous amount of growth on the other side. Which brings us to the engineers. Or rather, what counts as one. The categories of product manager, product designer, and engineer were always something of an artifact of how expensive it used to be to ship anything, but those lines are becoming blurrier by the minute.
No longer “how” but “what”
The questions that matter are no longer, “can we build this?” because we can build anything, but rather, “what should we build? What are we trying to accomplish? How will we know when it’s done, and how do we know if it solves the problem we defined in the first place?” Deciding what to do, for the foreseeable future, is a uniquely human endeavour. When the cost of building is cheap, the thinking work becomes the most important work: smaller groups, working closer to the problem, producing software that wouldn’t have been economical to write 6 months ago.
As a result, there’s a chorus shouting, “this is the end of work!” That AI is a bubble, and that all jobs are going to be taken away. This AI apocalypse narrative is bollocks, and we’ve seen it all before. The plow led to more farming, the typewriter led to more books, and Excel led to more work for accountants. Every technological expansion since the steam engine has added more roles and spawned new industries. For example, the global recruitment industry – which now approaches a trillion dollars annually – only started in your grandparent’s lifetime. The pattern has held every time. There is no serious reason to think AI will be the exception.
Beyond panic there is potential
Encouraging caution about transformative technology is a posture that speaks to fear and insecurity, and it does not engage with what’s actually happening: the surface area of work is about to expand again, and the new work, will be more creative, more human-to-human, and less administrative than what it replaces.
Builders vs. Architects
There’s also an uncomfortable truth we need to face. The differentiator for the software engineers behind the code is no longer the ability to produce code. It’s the thought-process that goes into what that code does, how different systems work together, and why it exists. Realistically, few people are good at deep thinking, and even fewer sit a problem for weeks, months, or years at a time. Now building is cheap, and we’re seeing software that solves “easy” problems getting eaten up every day. That means the premium on knowing what to build has gone up.
The importance of editing
The harder discipline isn’t producing more. It’s restraint. Knowing which problems to ignore, which features to refuse, which products are best left unbuilt. Cheap software means a flood of software, and most of it will be noise. The companies that thrive will be the ones that resist the temptation to build everything they technically can.
Choosing the harder problems that most people can’t solve on their own sits on every engineer picking work, every company picking engineers, and every society picking what to build with all this new capacity. The thinkers had better get serious.
