Build UNIX, not Uber

16 July 2022
16 Jul 2022
Seoul, South Korea
8 mins

I spent the last three weeks in Seoul, partly visiting family I hadn’t seen in years and partly taking a vacation from my usual work in my usual city in my usual context of life. While I haven’t been able to work much at all, I spent my subway rides and waiting-in-line times reading the Singularity Sky series of novels by Charles Stross.

These stories unfold in a post-singularity world with a multi-planetary human civilization. One of the central ideas that Stross explores in the series is the idea of economic scarcity. Is it necessary? What happens as technology chips away at it slowly but surely over time? What’s its relationship to human conflict? In a post-scarcity world, how does wealth accumulate? What meaning does money and wealth have?

It’s really an excellent series, and you should read it if you find any of these questions intriguing.

Exploring this post-scarcity universe (I’m still deep in the middle of it) made me question my personal conceptions of scarcity and how it influences the way we understand wealth, success, and the value of good ideas. It’s this relationship between economics and the value of ideas I want to invite you to explore today.

A few months ago, I had a chance to ask a certain famous founder of a multi-billion-dollar software company a question. I asked,

Silicon Valley likes to treat venture-backed corporations as a kind of universal hammer, approaching every large – even civilization-scale – problems by building companies. Given that you’ve contributed to venture-backed companies, nonprofits, and other unorthodox research initiatives, what do you think of “companies” as a way to solve large problems facing humanity?

He said he thought, despite the hype, companies were still an underrated tool for creating large-scale societal change. His reasoning: companies are uniquely independent in a world of bureaucracy, because companies are only ultimately beholden to customers.

There’s an element of truth in this. While being so independent and autonomous, good companies can amass huge pools of resources to coordinate large-scale action. This coordination power can be put to work exploring a wide breadth of scientific solutions, attempting huge construction projects, or motivating political change.

There are, of course, other ways humans coordinate power at massive scale. A lot of them look like religion (social movements, political activism, and of course, religion itself). Some of them increasingly look like software (open-source communities, trustless decentralized coordination built on cryptography). The rest of them look like markets (venture capital, taxes, science, global trade). It’s too early to tell what a mature ecosystem of software-defined coordination will look like. But between religion and markets, it’s probably easier for single human beings to build and scale coordination power with companies than with religious movements.

Companies amass coordination power by having good ideas, executing them using more good ideas, and capturing a big part of the resulting value created (that is, when they’re not extracting value out of monopoly controls over markets). The effectiveness of corporations that the founder alluded to above comes from two things being true:

In other words, companies cannot accumulate power if (1) everyone can get what they make for free, or (2) everyone can steal what they need or copy useful ideas without consequence. At various points in human history, these things has not been (or will not be) true. In a lawless society, companies can’t expect to capture most of the value they create – you get a world of piracy instead. In a hypothetical post-singularity society where ideas can be copied and designs can be reproduced for free in untraceable ways, companies can’t expect to capture most of the value they create. Companies, useful change-making instruments they may be, could turn out to be a gimmick of a particular phase of civilization – post-stability, pre-singularity.

A world in which companies can’t capture the worth of their work might sound apocalyptic, but I don’t think it has to be. There are corners of today’s world where we can observe what happens when companies don’t effectively capture the value of their good ideas.

The free and open-source software world is funded by companies, but not owned by corporations. Software is free to copy, and with the right permissive licenses, anyone with a good idea can copy and build upon existing open-source software to improve it or experiment with new ideas. If the stewards of a project are opposed to a new idea, proponents of the change can “fork” and try to win mindshare and support with their design or implementation.

Sometimes, companies lose ownership over their intellectual property by law. In 2024, the first versions of Mickey Mouse will enter public domain in the United States. Mickey will then join a pool of public domain literature and media that is free for anyone to remix, reuse, and reproduce for their own needs and creative tastes. This allows artists to create new work, obviously, but also allows initiatives like the free e-book library Project Gutenberg to exist. Project Gutenberg is one of several datasets used to train large language models like GPT-3. Free, public information has far-reaching consequences.

Elsewhere in the world, there are economic regimes that have intellectual property law that’s permissive or weakly enforced, to where ideas can be stolen and copied more easily. In China’s Shenzhen special economic zone, a lax approach to intellectual property law enforcement, in addition to the business and foreign investment-friendly policies in the zone, bloomed a tech hardware cornucopia. Shenzhen’s manufacturers and factories power a significant portion of the world’s consumer technology industry.

In all these environments where companies can’t capture the value of their work as effectively, the most powerful force in shaping society is how effectively good ideas spread in absence of explicit control and stewardship. Ideas that are useful, like material science and astronomy, spread effectively. But so do ideas that encourage their owners to spread them, like most religion; and ideas that grant their owners more power, like capitalism and weapons engineering. In a post-scarcity world, ideas are ascendant.

Of course, big ideas need resources like money and large-scale coordination to be implemented. Companies give their operators resources that’s useful for implementing ideas, like building rockets out of good rocket designs or reshaping democracy through an idea for how people can connect more easily online. But I think there’s reason to believe that resources required to implement big ideas are going down over time. Projects like the interstate highway system or the Manhattan Project required lots and lots of money and resources, but the Manhattan Project of today may be AGI, and AGI isn’t anywhere near as capital-intensive. More succinctly put:

Every 18 months, the minimum IQ necessary to destroy the world drops by one point.

There is reason to believe that, over the long term, humanity is moving from a regime where corporations steer society to a regime where autonomous ideas steer society.

We commonly view the history of tech invention and entrepreneurship through the lens of corporations and value capture. But I think there’s another way to study the twenty-first century tech ecosystem, as a battleground for good ideas, spreading autonomously.

UNIX was a good idea. It was free to spread widely across the tech ecosystem, and eventually some corporate players like Apple and Sun captured a lot of the value, but Linux stands out as the UNIX “winner” in 2022. Neither the core ideas of UNIX-style operating systems nor the Linux open-source project grew solely because of corporate stewardship. They spread and won because they were good ideas.

“Making inefficient markets more efficient with software” was a good idea, and companies like Uber, Doordash, and Etsy captured a lot of the value.

My favorite example is probably the Web. The Web – an interlinked collection of interactive documents made available through a decentralized Internet protocol – was a great idea. I think it’s no exaggeration to say that the Web won the platform wars between major operating system vendors without even really being an operating system. If you’re building a software company, 9 of 10 times, you better have a web application. The Web is an idea. No company owns it, and no company can.

I’m fascinated by this way of looking at how to make civilizational change: release good, powerful ideas into the world. And if circumstances allow, try to capture some of the value that comes from those ideas shaping the world in their image. Perhaps corporations will persist as the hammer and chisel with which history is carved through our generation. But perhaps not.

Charles Stross writes of Manfred, the protagonist in his novel Accelerando:

“You are very unusual. You earn no money, do you? But you are rich, because grateful people who have benefited from your work give you everything you need. You are like a medieval troubadour who has found favor with the aristocracy. Your labor is not alienated – it is given freely, and your means of production is with you always, inside your head.”

and later,

“You want to abolish scarcity, not just money!”

“Indeed.” Gianni grins. “There’s more to that than mere economic performance; you have to consider abundance as a factor. Don’t plan the economy; take things out of the economy. Do you pay for the air you breathe? Should uploaded minds – who will be the backbone of our economy, by and by – have to pay for processor cycles? No and no.

As the technology machine chips away at scarcity with its power tool called software, I wonder if our most resilient legacies would be made by searching for ideas that spread more than companies that scale. Building UNIX, not Uber.

Design with materials, not features

I share new posts on my newsletter. If you liked this one, you should consider joining the list.

Have a comment or response? You can email me.