For the past 15 years I have been writing open source software to fight the downsides of capitalism, and then the watershed happened. The cost of producing software has been in a free fall over the last months. This has an impact on my work in three places.
First, proprietary software will become cheaper and more replaceable, partially removing the problem that open source fixes. Companies like SAP, Oracle and Microsoft are no longer the only ones who can produce the high-quality software that enterprise and government organisations need, so their power wil wane. Instead the power will go to AI companies, so arguably open source funding should go to projects in self-hostable open source LLMs, rather than in data portability.
Second, open source software will become better with nicer UI and faster bug fixes. These are exciting times, where we can just point at for instance a decade-old html app and tell Claude to "modernize the code base", "add tests", or "move the server-side code into the browser". This is what I have been doing this week in the new AI-powered incarnation of Ponder Source.
And third, the process of developing open source alternatives for proprietary software will become easier to organise. We no longer need big teams and funding, we just need a requirements specification, 20 dollars for a Claude subscription, and then the software just appears, like a plastic-wrapped meal from a microwave.
Sure, as Cory Doctorow and Rainer Mühlhoff argue, the current boom of the AI industry is becoming a weapon of the super rich. Ten years ago we were worried that facebook was eating up the open web. Now it might be more than just the open web that is being eaten - the whole knowledge economy is under threat of being replaced by AI. Democracy and regulation has always stood in the way of industry, but if industry develops cognitive powers, including the ability to earn money more easily than humans do, and to persuade people into thinking and voting in its favour, then regulation becomes harder and harder to maintain. I don't know how to fix that.
And while industry is becoming more powerful in its ability to go against the public interest, individual humans are losing an important power they have had over the past two centuries or so: the ability to work for money. Capitalism is always defended as a fair system for humans, because everybody can work for money. That is about to change. A year ago I could still earn 100 dollars per hour just by writing TypeScript code and Dockerfiles. This year, the job market for programmers is simply gone. In a year from now, the same will be true for not just the spreadsheet workers but also the lower level managers at for instance insurance companies.
At first, the office buildings will not immediately be emptied out, because the drop in price will cause an increase in activity, not a decrease, but eventually I think they will be replaced by a few big data centers, and many humans will lose their ability to work for money. The industrial revolution took away a lot of unschooled jobs. In the coming decade or so, the AI revolution will also take way many of the schooled ones.
It is hard to predict what will happen then. I think people will still keep on working regardless, we are good at finding things to do. But eventually, in a hundred years from now, the monetary economy is likely to be taken over by autonomous artificial proprietors.
So while I will keep following the debate about Collaborative Finance, Commoning, and the post-capitalist future of the human economy, I will focus my work on using AI to build the open source tools we have always been dreaming of.