The Tame Impala-fication of Software Development

Let me tell you something about how software gets built these days. It used to require armies. You needed designers, backend engineers, frontend engineers, DevOps people, QA, marketing, the whole circus. And sure, for certain classes of problems that still holds. Building Google’s search engine? You’re going to need more than one person. But building a product that solves a real problem for real people? That landscape has fundamentally shifted.

Kevin Parker (Tame Impala) made Elephant, (one of my all time favorite psych rock songs) all by himself. Not partially – entirely. Wrote it, performed it, recorded it, produced it. One person with a vision. And the result was a bloody chart-topper. No committee, no label-mandated co-writers, no producer’s committee editing his work. Just one guy in a room making something happen. Now look at software development. Twenty years ago, if you wanted to ship a product you needed a team. You’d need someone who understood databases, someone who could make it look good in a browser, someone who could figure out how to get it running on servers, and someone who could convince people to actually use the thing. That’s four minimum, and those four people needed to coordinate, which meant meetings, which meant overhead, which meant things moving slowly.

Today? One person can do all of that. And I don’t mean “one person manages a team of contractors” – I mean one person, sitting at one computer, building something that works. Everything has changed:

The tools got democratized. You don’t need to know how to configure a Linux server anymore. AWS, Vercel, Railway, Fly.io – pick your poison, deploy with a few commands. Stripe handles payments. Supabase gives you a database with an API. The plumbing is solved, and the plumbing used to be half the battle.

The distribution is free. Remember when you needed a sales team to get software in front of people? Now there’s Twitter, Product Hunt, Indie Hackers, Reddit, TikTok. You can reach your exact audience without spending a single pound on advertising. The bottleneck shifted from “how do I find users” to “do I have something worth finding.”

AI became a co-pilot, not just a buzzword. I’m not saying AI writes your code (sometimes it does, and it’s bloody useful). I’m saying it fills the gaps. Not sure how to implement auth? AI’s got your back. Need to write some SQL? AI’s got your back. Need to draft marketing copy? AI’s got your back. The scope of what one person can effectively attempt has expanded dramatically.

And this is where it gets interesting. Because the question isn’t “can one person build software” anymore – they clearly can. The question is: what does one person make?

Sure, Kevin had access to good instruments, but he also had a vision. He knew what he wanted the music to feel like. The tools let him execute, but the vision came from somewhere else entirely.

Same with software. The infrastructure is commoditized. The patterns are well-documented. The templates exist. What remains is the hard part: knowing what to build, knowing who it’s for, and having the taste to make something people actually want to use.

This is why I think we’re going to see a wave of incredibly capable solo developers who ship things that used to require teams, simply because the game changed. The floor dropped out from under the pre-requisites, and what’s left is the part that’s argubly always mattered most: the ability to see something that doesn’t exist yet and figure out how to make it real (very Steve Jobs, I know).

So here’s my question for you: what do you want to make? Because the excuses are gone. The tools are now here. The distribution is here. The only thing missing is someone willing to sit down and do the work.

Go make something.