Keep it simple, stupid.

It's 2024. We've got over 700 core programming languages and hundreds of layers of abstraction, which all ultimately boils down to the same kind of machine code that Kathleen Booth created in the 1950s. Back then, every single bit and byte was carefully counted for, ever watt of power used and no capacitor wasted.

Today, it's no secret that tools like blockchain and generative AI can use so much computing power they outstrip some small nations in terms of energy consumption. Have we overcomplicated our tech to the point of relative absurdity, or is this just the price we pay for progress?

They Don't Make 'em Like They Used To

I was inspired yesterday by the story about NASA's Voyager 1 probe. Last year the space vehicle, which launched fourty-six-and-a-half years ago and is still travelling away from Earth at a speed of over 17 kilometers per second, experienced its first major computer failure.

Let me rephrase. The computer onboard Voyager 1 had been running continuously for 16,900 days before experiencing its first major computer crash. I don't know about you, but my smartphone barely manages sixteen days without needing a restart.

Now that's reliability!

Apples and Spaceships

Comparing the computer systems of Voyager 1 to a smartphone is a bit like comparing apples to... well... spaceships. My phone can do hundreds of different things, many of them at once. It's a complex array of sensors and radios and compatibility layers that allow me to throw a more computing tasks at it than a computer scientist could have dreamed of in 1977.

Voyager, by contrast, does fewer things better. It also has a complex array of sensors and radios - but the use cases for these are tiny. The probe can take hugely advanced spectral readings of a planet's surface and transmit them through a deep space network. Sure. But can it be reprogrammed to do some complex analysis of that data, tweet it, and then monitor the replies? No, sadly not.

It's taken a few months, but NASA have managed to repair Voyager 1 by patching its software from over 24 billion kilometers away. Try to get your head around that. A computer system older than half the people reading this was updated and repaired remotely by a tiny team who had no physical access to it, and who largely hadn't been alive when it was created.

The New Frontier

Back in the 1990s I became obsessed with the demoscene. I loved computer generated music - but more than that, I loved the idea of trying to fit something seemingly complex and beautiful onto a single floppy disk. Every byte mattered.

Recently, I've been thinking a lot about how this has changed. Fast internet and cheap storage means it's not uncommon to be sharing terrabytes of data between teams on an average week. It's no wonder we've stopped caring about data and power usage when the world enables so much more.

But honestly? I'm starting to think we need to return to the idea of reliability, scalability, and anti-complexity. Let's build tools and services that don't fall over, that foster trust, and that are easy to extend because they aren't behmoths of spaghetti code.

And so, as Voyager 1 reconnects with planet Earth on its continuing journey out of our solar system, I feel inspired to create equally interstellar feats of engineering whilst squeezing the most out of every watt, byte, and line of code.

And who knows, perhaps we can create the planet's next pale blue dot moment.