I wasn’t independently wealthy. I didn’t have a trust fund. I just wanted to build something good. How could anyone like me, with experience but without a financial safety net, possibly win?
Ben Werdmuller
Suppose you didn’t want to build a web platform, rather you wanted to fly a rocket to the moon. How could a single individual possibly achieve that if there’s no way to funnel enormous amounts of money through that individual?
The answer is, a single individual can’t achieve flying a rocket to the moon. In fact, they should never be permitted to even start trying, because they are likely to damage themselves and the people around them.
I believe the technology industry is flailing so badly because of its “startup” culture. It mythologises a hero’s journey of a genius individual struck with vision, who begins tinkering on their own, and gradually builds up an enormous pyramid of millions of followers beneath him†. Our patriarchal hierarchical society cannot imagine organisations shaped in any way other than a pyramid with a single individual at the top. But our protestant christian culture cannot imagine a moral justification for anyone occupying the top of that pyramid except by working hard. The solution is to imagine the sole founder starting on his own, without any assistance. As the prime mover he is morally justified in owning whatever gets built by the people below him. Yes, some subordinate engineer may have slaved away in a cubicle for 120 hours a week to produce a miracle of technology, but they wouldn’t even have started without the founder. So it makes sense for the founder to take the credit. Right?
This is actually sorta kinda plausible with technology, because all you really need is a laptop and lots of time to concentrate. It’s clearly insane with moon rockets. And yet, the psychological pull is so powerful that we are now trying to force the founder model onto the moon rocket industry as well. Only centi-billionaires are eligible to become such founders, but it’s impolite to dwell on that.
However, it turns out that in practice it’s not plausible with technology either. You can produce a working piece of software on your own. But you cannot produce a good piece of software on your own. Producing good software does, in fact, require just as many talented people working together according to complex and strict rules of governance as building moon rockets. Meanwhile, the effort required to produce non-good working software is so enormous that thousands of people are destroying themselves every day trying to become Linus Torvalds. The best results constitute the current landscape of open source projects. The worst results are the quiet tragedies that we all avoid discussing.
There is an alternative way. It’s called democracy. A group of people with a common public interest votes for an executive representative council. The executive levies taxes on the members of the group, and determines which projects receive funding in order to achieve that common public interest. And at the same time the executive establishes subsidiary, quasi-autonomous governance structures to ensure that the funding is used efficiently to achieve the particular goal the executive assigned. No individual is critical to this process, the governance frameworks allow substituting new individuals as required. The quasi-autonomy ensures institutions are somewhat robust to changing priorities, ensuring continuity. But the overall system is still fundamentally driven by the common public interest, and responsive to how that changes.
In a way, the question “how can we ensure that gifted founders have the money they need to realise their visions?” is itself an attack on democracy. Alternative models such as venture capital are all predicated on the idea that the gifted founder owns the pyramid that grows beneath him, and is entitled to sell that property as he sees fit. Whereas under democracy, the prime mover of a new technology is simply an employee doing their job, with no special claim of ownership. Democracy implies a circular chain of accountability, with the heads of organisations ultimately answerable to the same people who are the “feet” of those organisations. Any structure that leaves a particular genius visionary in a privileged position of power is incompatible with democracy.
If you are an engineer who believes that the open social web is important and powerful and practically achievable, and has an idea of exactly how it could be realised given a little bit of funding, then the democratic approach for you is to pitch your idea to your local democratic institutions. You would work to gather a coalition of other employees with the necessary skills. Together the group would either establish a new institution with the necessary accountability and governance structures, or redirect the efforts of an existing institution. And as an individual you would become just one cog in this new institutional machine, cooperating to make the world a better place. If you later change jobs, the machine keeps turning.
The problem Ben Werdmuller is wrestling with here is, but but but, there just is no democratic institution that could even begin working on the open social web.
That is true. That is because he’s talking about software. While other industries have a more obvious history and culture of state direction, the software industry has a mindset entirely dominated by this myth of founders working outside of, often in flat-out conflict against, “the government”‡. It’s almost inconceivable to re-establish the software industry on a democratic foundation. You need to persuade the people with authority, and in a democracy that is not the politicians, it’s the electorate. “Hey, great idea everyone, let’s increase taxes and give that money to highly-paid nerds!” Changing the mindset of an entire country is an enormous project. And as someone raised on the myth of individual people achieving great things starting from small beginnings, Ben Werdmuller shies away from big up-front solutions like this.
Nevertheless, we can make progress here. Perhaps we can’t skip straight to “I am writing code for my open social web platform”. My pitch instead would be a new institution dedicated to improving government IT systems from both sides at once: how people work with software, and how the software supports their work. I would come up with some snappy title for this, let’s say “department of government efficiency”. This would employ the usual collection of specialists: product managers, business analysts, market researchers, testers, customer relations, data scientists, and engineers. These would start from the ground up fielding day-to-day problem reports from government IT users. They wouldn’t start with a particular software product in mind, they would touch whatever software their customers need to use. They would work to solve the problems on a small scale, both by educating users and by filing the rough edges off the software they use. This would be expensive, but predictable and flexible, and it would immediately produce visible and measurable improvements in the efficiency of everyone else in government.
This is a genuine practical proposal. Similar institutions are created pretty frequently by medium to large governments, even in sclerotic, technology-hostile countries like the USA. I don’t know how these ideas come to fruition, but I imagine that they start with knots of budget-constrained mid-level managers sharing the frustrations of being unable to tackle the obvious roots of their problems. And this idea doesn’t need to start big. Hiring a team of fifty or a hundred professionals comes with a high price tag, but so do the unusable technology boondoggles with which governments are routinely infested. DOGE could justify its existence by salvaging just one of these.
A natural byproduct of DOGE’s work would be ideas for more ambitious software projects with deeper consequences. But it is only through the daily grind of solving small problems that the institution can gain the knowledge of how to prove that software work pays for itself. They can then apply this methodology to bigger projects, presenting detailed cost/benefit analyses to higher managers with larger budgets.
Eventually, perhaps one of the projects DOGE might tackle is helping the government communicate with the general public via open social web platforms. But honestly, I can’t see that happening for very many years. That’s because that is not the most practically useful application of our resources. And that is why Ben Werdmuller is having so much trouble getting funding.
The key is to start gently and quietly, doing difficult, rewarding, but unspectacular work to help other people do their jobs better. This is the way to start from small beginnings and scale up. But while it might start small in one way, for an individual technologist it still requires a radical revolution in thinking. It requires them to discard the myth of the genius founder, and the habit of resorting to for-profit venture capital needs to go with it. Those things got us where we are today.
Instead of seeing ourselves as individual actors, we should place our actions in the context of a wider, unavoidably political society. And I fear that no-one with a history of working in “startups” is capable of making that leap. It is a trap to believe that because any kind of business can begin with private funding sources, private funding sources can achieve any goal at all. The reason they don’t produce the results we want is not that no-one ever thought of pointing them in the right direction. It’s that what they do now is all that they can ever do.
† Pronoun chosen advisedly.
‡ A useful habit when consuming modern political discourse, is whenever someone says the word “government”, mentally replace that with the word “democracy”. “We want to pursue a small-goverment approach”. “Government has no business interfering in our lives”. “Government is always the problem, never the solution”. It really helps clarify the agenda.