DIY build log: Human society v1.0.0

Earth here, at it again.

I've been avoiding this for a hot minute – bricked too many. (You should see the fumes coming off my last one.) It's tough because you can never really start fresh. You fork a few, patch holes, and pray.

That said, humming along on my desk here, I have a prototype I'm pretty happy with. Here's how I spun it up.

Parts

• 1 × Ecosystem
• 1 × Industrial supply chain
• 1 × Constitution
• 1 × Tax bureau
• N × Fallible primates

1. Write the smallest kernel you can

Rome, one of my grander attempts, didn’t fall in a day. It rusted to a halt. Eventually the Visigoths stripped it for parts. Too much territory, military overhead, all that jazz. Elite capture ran amok, citizens checked out, and no amount of new taxation or slave labor was ever going to fix it.

Cutting scope like a madman, as expected, worked wonders after that. Short constitutions, explicit rights, clear amendment pathways, no permanent offices. As you'll see, other problems arose, but this was huge for me.

Fear still wants a sovereign, so I always install one, but with a ton of proxies around it. Split it up. Branch early. Power also wants to accumulate, but if you put it through a well-crafted little maze, it grows better – less like a hydra, more like an irrigated field that people benefit from. Participate in.

Your core files, in a nutshell, cannot require a priestly class to interpret them. If so, you’ve basically lost already. Write them like you might spawn in your society as the worst-off, poorest, most ignorant person in it. Keep it simple, stupid. Some people will be. That's alright.

2. Bootstrap revenue and morality together

Political centralization is one beast, economic is another. It'll eventually hoover up any government you give it.

One of my modules, America, wouldn't stop running into this. After Rome I was hellbent on being hands-off at the beginning – separation of concerns! – so I wrote some of the best code in my career. Its alignment was utter trash. Civil war, less than a century in. When your revenue model conflicts with your stated values, you're deferring bloodshed, and I was deferring a lot with that one.

It heralded a wave of similar bugs. Colonial atrocities. World wars. Things got bad. When computing entered the picture, the extraction morphed; the hoarding ended up worse than before; people in the imperial cores had a hard time resisting. The extraction, running out of places to go, had wormed its way into their minds!

A steaming pile of imperial boomerangs before me, I wondered: how do you avoid that? Or reduce the surface area for it? The thing about power’s corrupting influence is that, while it can only go so far with thrones, it scales like nobody's business with institutions. It spawns all kinds of processes. Bentham and his panopticon. Klein and her "shock doctrine."

Looking at isolated pockets in my last model where people avoided this (it was even working in the biggest American city right toward the end), I managed to come up with a fix. Economic and moral (or "political," or "social," whatever) concerns need to be tied on the smallest levels. People need a path to building and maintaining the resources they need locally, which creates immunity to endless extraction by global megacorps. Infrastructural, collective, grassroots stuff.

Stuff like land and health gets prioritized and protected; systemic sins like hoarding and addictive products get taxed and stay taxed. Jury trials and trust-busting keep their bite, campaign finance gets bitten. Budgets are transparent and people want to review them. Security flaws shrivel under many eyeballs. The economy doesn't drift so easily into moral contradiction.

3. Firewall the public info layer

Dictating those things, all that morality stuff, consistently fails. Humans are restless and fiddly. They change their minds. They talk; tell stories, deliberate, hash things out. This means the public information layer is the first and ugliest thing to fall when a group of them starts getting all extract-y on you, which is hard to avoid entirely. Immediately borks most of the safeguards mentioned above.

Regimes lose the people's belief, and their own institutional legitimacy, way before they lose territory. On the flipside, some societies in the early Americas landed on a lot of the exact solutions I've talked about here, having done so much political experimenting; incoming societies didn't want to hear it. Called them savages, practically wiped them out. Propaganda spreads, and sticks, like little else.

Independent journalism: be sure to have it well-funded, well-insulated.
 Algorithmic transparency: be sure to have it enforced. Freedom-of-information requests should be, like, a blink away. Maybe most of all: make education teach argument and investigation, not pledges and slogans. Humans are naturally curious; unfortunately, there are ways – and I saw a lot of ruling classes do this with varying success – to make them unnaturally ignorant.

Undefined behavior makes the world interesting. Without shared reality, undefined behavior is all you get. Runtime errors abound.

4. Give nature root privileges

Humans are good at many things. Even the most ecologically conscious among them still can't know all the second-/third-/fourth-order effects they have on the web of life.

I watched a bunch of cultures avoid this simply by never extracting so much that those effects become a problem. Some saw those problems here and there. A few ruined themselves with those problems.

Even when they're otherwise equitable, getting along fine, not lost in insane stories, this can happen. They'll agree on a course of action that just doesn't work with where they live: overfishing, overlogging, overhunting. Lead, fossil fuels, microplastics, processed food, short-form video. I've seen the gamut.

The definition of "nature" is broad: they can extract too much from each other's minds, too. They can get too stressed, brain-rotted, complacent – without breaking the public info layer! – to use those perfect political processes you've built to fix the environmental problems.

They get all uppity (and/or simply ignorant) about how what "masters" of nature they are, develop a habit of boundless extraction, and melt your machine in no time flat.

And when the environment crashes – maybe not even that much, maybe just enough for them to panic – all your hard work can go out the window in their frenzy to restore normality. They take shortcuts. Nothing in their books or toolbelts prepared them.

In a nutshell: some pollution fines need to be unavoidable. Some biodiversity zones need to be untouchable. The whole economy has to be as cyclical as physically possible. Low waste. Nuclear energy. Regenerative agriculture out the wazoo. And it all needs to be like that before the worst stuff happens.

Sometimes there's just no room for “we’ll fix it later.” Climate is a constraint, not a debate. I've tried negotiating with thermodynamics before. Never seems to work for me. If you have workarounds, drop me a line.

Remaining bugs

• Power reconcentrates. More slowly but it tries. Wealth pools in new ways, bureaucracies defend themselves, emergencies tempt shortcuts. Anywhere there’s a hint of structurelessness, tyranny rears its head. You have to stay creative about giving it friction, patching up holes.
• Citizens get tired. Participation gets difficult as complexity grows. Accurate information isn’t "fun" unless you make it so. Focus less on ideals here than on preventing failure modes. Good local governance is a bedrock, a key piece of defensive architecture. Make people take turns. Do that well, and buy-in will continue, trains will keep running, bread and circuses will see innovation without flying off the handle. Incentives will wobble, but stay aligned. Iterate upward and see results.

It’s not utopia (never will be) but so far, it keeps on booting. Handles stress better than previous builds, I'll tell you that much. Fails slower, corrects faster, doesn’t gobble up its own future. Seems like it should stick around awhile.

Will provide an update in a few centuries on how it's doing.