The Home-cooked Computer Club
Perhaps I’m getting a little too old. A little too like the person you avoid since they are always going on about the “good old days”.
When I first started writing software, there was a clear divide-hobbyists, and professionals. Professionals wrote for serious computers (Mainframes and Minicomputers) using series languages like COBOL. Hobbyists were spawned by the emergence of Microcomputers, what we came to call PCs in the middle to late 1970s. Hobbyists used languages derided then and now, particularly BASIC (which as we noted recently turns 60 this year). And programmed computers derided then as toys. Oh yeah in the industry there has always been a lot of gatekeeping.
But a strange thing happened. The derided toy computers and programming languages, and the hobbyists that created and used them, transformed computing. What had been the preserve of deep pocketed corporations and governments and large Universities became something any relatively middle class western family could afford.
Microsoft was founded by essentially a hobbyist, Bill Gates (with Paul Allen) making software (originally BASIC) for hobbyists and microcomputers (hence the name). And Apple (Steve Wozniak and Jobs were involved with the Homebrew Computer Club), a group in the Bay Area that included other early microcomputer inventors like Adam Osbourne.
Some of these hobbyists went on to become immensely powerful, and wealthy. And their products to transform the world economy.
Fewer of the enthusiastic hobbyists I knew as an early teenager did. The Doctors and Accountants and middle class professionals who solely out of fascination bought these new devices, met in community halls, formed tribes associated with their commuter of choice (the mild-mannered accountant whose kids I babysat and his Apricot owning friends would not mention the word Apple in relation to computers–it was always “the other fruit named computer”).
Why was this moment transformative? It made computing vastly more available and vastly cheaper. And so it made programming more accessible. Sure to mostly western, middle class, white men, but it still increased the number of people who could program a computer by orders of magnitude. And the cost of software development similarly.
A single person or very small team could write world changing software. Dan Bricklin and Bob Franklin wrote VisiCalc, the original spreadsheet in 2 months for the Apple ||. These days setting up a new web project might take 2 months! (Ok I jest, a little).
But little now doesn’t seem to be built without large teams and (as a consequence) require venture capital, or within existing companies. Perhaps the hobbyist was simply an artefact of a moment in history, one professionalised out of existence.
So where is this going John? Well two things I read this week have prompted these musings.
First, a new talk by Maggie Appleton, who delivered an amazing keynote at Web Directions Summit last year. Watch that here (no signin or singup required). In Home-Cooked Software and Barefoot Developers Maggie observes
For the last ~year I’ve been keeping a close eye on how language models capabilities meaningfully change the speed, ease, and accessibility of software development. The slightly bold theory I put forward in this talk is that we’re on a verge of a golden age of local, home-cooked software and a new kind of developer – what I’ve called the barefoot developer.
Home-cooked is the new home-brew. Maggie comes from a position in opposition to the modern, VC driven software world (and one that aligns with my recent Project Dissenshittify. But what was really interesting is the following day I came across this, The End of Software, from a VC, Chris Paik.
Software is expensive because developers are expensive. They are skilled translators–they translate human language into computer language and vice-versa. LLMs have proven themselves to be remarkably efficient at this and will drive the cost of creating software to zero. What happens when software no longer has to make money? We will experience a Cambrian explosion of software, the same way we did with content.
Coming from two very different starting points, two smart accomplished people ask similar questions.
Not all, indeed, not many, toys grow up to transform the world. The ones that do, do so by transforming economics. Making something orders of magnitude cheaper or easier is not much different from enabling entirely new things.
The Greeks before the common era had steam engines like the aeolipile. But it was likely no more than an entertainment or toy. Nearly 2 millennia later, steam power, specifically Watt’s first effective steam engine, by disconnecting the output of human or animal labour from economic productivity made energy orders of magnitude less expensive, and ignited the Industrial Revolution.
What impact on the economics of software development will large language models have? From a coupIe of years of intensive personal experience I can attest to them making me much more productive, not simply allowing me to do what I have long done more quickly, but to attempt things I know I could do, if absolutely necessary, if I had an unspecified amount of time, but which I baulked at not knowing how much time or effort it might require. Like exploring APIs and even programming languages I’m almost entirely unfamiliar.
My experience too has shown me that these technologies don’t devalue the deeper knowledge and intuitions I’ve developed since those days of BASIC and microcomputers. Well not yet. And my instincts here are that we are quite some way yet from folks without software engineering knowledge being able to develop secure performant, usable software with LLM based tools. Just as we are from LLMs replacing doctors and others with complex domain expertise. The day may come, but I’d hazard it’s a way off yet.
But acquiring those capabilities and intuitions will also be accelerated by these technologies. They can explain code to you. They can help you learn about aspects of a programming language you may not be familiar with. They can help you improve the quality of your code. Are they right all the time about everything? No. That’s one important reason why those intuitions, that deep knowledge is important.
My instincts are too that Maggie Appleton and Chris Paik, coming from two very different directions are onto something. That the economics of software development will change. Problems that haven’t made economical sense to solve with a VC funding model make economic sense when the cost of software development falls by a factor of ten or a hundred.
Web Directions Next
If you think ideas like this are worth exploring, our brand new conference Web Directions Next, a one day single track conference in Sydney November 29th is created to explore these sort of ideas-the impacts and implications of changes in technology in the broadest sense. We’d love to see you there.
Great reading, every weekend.
We round up the best writing about the web and send it your way each Friday.