When people talk about politics, they generally fall into one of three groups.
The first and by far the largest group basically argues “more good stuff, less bad stuff!” They may use clever rhetoric, but they fundamentally ignore the idea of tradeoffs. They view policy debates as a parade of applause lights or boo lights, and cheer for anything that sounds vaguely nice. They can be persuaded to support or oppose any policy just by changing the wording from “provide more government services” to “raise taxes to make big government bigger.” All Moldbuggian arguments aside, this group of people is the best possible argument against democracy.
The second group takes a more sophisticated route. They understand that most policy decisions involve trade-offs, and argue accordingly. They may claim that yes, government welfare does distort incentives somewhat, but at the margin we should be doing a little bit more to help the poorest people in society. Or perhaps aggressive policing does represent an encroachment on civil liberties, but that public safety is important enough that we should empower the police a little bit more. These sorts of debates can get just as heated as the first sort, but they’re noticeably saner. Although these people may disagree on values, everyone is aware that they’re working off a reasonable and sane description of how the world actually works.
The third group takes things to an even higher level of nuance and sophistication. They basically argue “more good stuff, less bad stuff!”
I had a hard time really understanding the neoreactionary term “anarcho-tyranny.” Wikipedia credits it to Samuel T Francis, who had this to say about it:
What we have in this country today, then, is both anarchy (the failure of the state to enforce the laws) and, at the same time, tyranny – the enforcement of laws by the state for oppressive purposes; the criminalization of the law-abiding and innocent through exorbitant taxation, bureaucratic regulation, the invasion of privacy, and the engineering of social institutions, such as the family and local schools; the imposition of thought control through “sensitivity training” and multiculturalist curricula, “hate crime” laws, gun-control laws that punish or disarm otherwise law-abiding citizens but have no impact on violent criminals who get guns illegally, and a vast labyrinth of other measures. In a word, anarcho-tyranny.
The problem was that it seemed like an empty piece of rhetoric. You could point out any bit of degeneracy and call it anarchy, then turn around and point to any policy you dislike as tyranny. So all “anarcho-tyranny” boils down to is “hey, the world doesn’t totally conform to my standards!” – which for anyone of any ideological stripe is always a pretty good bet.
But I’ve come to understand that anarcho-tyranny, properly used, is a much more specific diagnosis. It is describing a particular set of adaptations that societies are forced to use when they’re living in the aftermath of a collapse in what I’ll call social technology.
Group 2, the one that argues politics in terms of tradeoffs, is basically correct about how policy works. Governments decide how much tax money to raise and give to the poor, and how strong their police forces need to be to control crime – keeping in mind that both taxes and police have downsides and it’s quite possible to have too much of either. You could view this as a production possibilities frontier, with a standard exchange rate between taxes raised and poor people helped, or between police power and deterred crime. And you can ponder this curve, look at where we’re standing at a society, and argue about whether we should be sliding a little bit up or a little bit down this curve.
What this tug-of-war game misses is that it’s sometimes possible to get better at both sides of the bargain. When medieval charities organized early hospitals in Europe, the economies of scale meant that you could actually help more poor sick people per dollar raised. When the police are able to work effectively with a law-abiding citizenry, they can be more effective at stopping crime while still being Officer Friendly.
We use the term “technology” when we discover a process that lets you get more output for less investment, whether you’re trying to produce gallons of oil or terabytes of storage. We need a term for this kind of institutional metis – a way to get more social good for every social sacrifice you have to make – and “social technology” fits the bill. Along with the more conventional sort of technology, it has led to most of the good things that we enjoy today. 
The flip side, of course, is that when you lose social technology, both sides of the bargain get worse. You keep raising taxes yet the lot of the poor still deteriorates. You spend tons of money on prisons and have a militarized police force, yet they seem unable to stop muggings and murder. And this is the double bind that “anarcho-tyranny” addresses. Once you start losing social technology, you’re forced into really unpleasant tradeoffs, where you have sacrifice along two axes of things you really value. 
Start with a pleasant town, with a trusting, cooperative, watched over by professional, friendly policemen who knows the streets, and the townsfolk, like the back of their hands. Crime is rare, but when it happens, it’s news, and the people, from the concerned neighbor to the kid who reads way too many detective novels, eagerly overwhelm the police with their offers of help. With that information and cooperation, what crimes there are usually are solved without so much as a baton broken out, and the community is grateful for being protected.
Then, almost imperceptibly, things begin to change. There begins to grow an idea that authority figures are not to be trusted. At first it’s just a few kids at the local college with some odd political ideas, but pretty soon the town’s poor pick up on it too, with the notion that their loyalty is to their fellow poor rather than to the town’s government and police force. And before long, even otherwise law-abiding people are refusing to cooperate with any police investigations. At the same time, riding on a wave of new complaints about police oppression, the state starts passing laws that make it a harder and longer process to hand down long prison sentences. The police is backed into a corner, both less capable of gathering information to investigate crimes, and facing an uphill legal battle to prosecute alleged criminals. Crime starts creeping up, despite the promises of increasingly desperate mayors and commissioners. Being a beat cop looks like a terrible career choice, the smart kids stay away and so the ranks start getting filled with less qualified candidates, who care less about their work and more about their pensions. The force becomes increasingly alienated from the town, seeing the poor districts as enemy territory, and it starts seeming less bad to use a little enhanced interrogation to crack a tough case.
You graduated from the local college and settled down for the long haul. It’s election season, and you flip to the mayoral debate. Police reform is the hot button issue, and it’s something that’s been on your mind a lot lately. One of your friends had his car broken into a month ago. Just a few weeks ago you heard that one of your coworkers got mugged and you do a double take – that’s a street that you used to play on as a kid. Naturally, you resent the spike in crime. But at the same time, you’re disgusted at the increasingly boorish behavior of the police – it seems like there’s always a police brutality scandal in the news, and you know that’s only the ones you hear about. The moderator finishes his question and one candidate responds, saying that the only way to get a handle on crime is to give more and more powers to the police forces – which, a little voice whispers, you know they’ll just end up abusing. His opponent responds, saying that the real problem is the growing militarization of the police, leading to the citizenry understandably unwilling to work with them. The only way forward is to assuage fear of police brutality by giving more protections to those accused of a crime – many of which, the little voice whispers, are in fact guilty. No matter what you do, all that either candidate can offer is with higher crime and a more unpleasant justice system than you remember from your childhood. You turn off the television in disgust, and decide that this November you’re going to do the one thing that will really make a difference. You’re going to write a strongly worded letter to the editor. 
The fascinating thing about this notion of social technology is that once you start looking at politics through this lens, you start seeing its influence everywhere. A lot of our thorniest political battles involve two sides desperately trying not to lose too much ground in the aftermath of a loss of social technology. Civil libertarians bemoaned the growth of the police state while law and order types bewail soaring crime statistics. Left-wingers bemoan the lack of progress on poverty, while right-wingers point out that the ratchet of federal funding is clicking upwards with no end in sight. And because of human loss aversion, these fights are often some of the most bitter debates in politics. Both sides remember a time when they were much closer to achieving their values, and neither wants to backslide further now.
So that is the value of the term anarcho-tyranny. It points out these situations where we as a society are getting less for more – where we’re losing social technology. Of course, just because we can diagnose a loss of social technology doesn’t mean it’s straightforward to figure out exactly what delicate institutions were giving us the bounty to begin with. And it’s even trickier to figure out what we need to do so that our institutions have incentives to be more and more functional – to sustain the same kind of progress in social technology that we have experienced in material technology. Politics is hard, and the idea of social technology doesn’t suddenly make everything easier. But it points us in the right direction, and starts us asking more interesting questions.
The third level of political discussion is “what happened to our social technology, and how do we get it back?”
 In fact, this is more or less explicitly the argument that Palantir made as to why they were not evil, despite running essentially a huge intelligence bureau. They argued that the government was going to do whatever it took to prevent another terrorist attack, no matter how intrusive it would have to become. So the only way that directive wouldn’t result in Orwellian levels of domestic spying is if they had access to software that let them more efficiently target the bad guys without needing as much invasive snooping on ordinary citizens – if they got better technology. Now, given recent revelations about the extent of CIA snooping, we may well laugh at that argument, but the Palantir folks would likely argue, and with some justification, that without their tools, domestic surveillance today would be even worse.
 I am aware that there is a second, distinct use of the term “anarcho-tyranny” – that when the government deliberately tilts the playing field of a tribal conflict by letting Group A terrorize Group B, while cracking down tyrannically when Group B attempts to defend itself or retaliate. This is a real phenomenon – whether in Nazi-Communist street fights in Weimar Germany or in the coverup in Rotherham. But having one name for two concepts does justice to neither, so I will use anarcho-tyranny only in the context of losing social technology, and leave others to come up with a suitable term for the second concept.
 This, in macrocosm, is basically the story of crime in the 20th century. Crime and policing both started out at a relatively benign level. A decline in social technology – involving some of the factors in the story – led to a spike in crime that peaked in 1970. Since then, politicians and police forces have been painstakingly fighting back by massively increasing the budgets and powers of the justice system.
Some people have argued against the idea that crime has gotten much much worse over this time period, on the basis that crime stats have mostly stabilized and even improved since the 1970s. What this argument misses is the enormous investments that have made this relative improvement possible. The police have gotten more expansive powers and more sophisticated technology, the level of incarceration in the US has reached staggering levels, and any private citizen who can afford it spend huge amounts of money cocooning themselves. They pay exorbitant rents to live in safe neighborhoods within cities (where large areas are effectively uninhabitable), move out to the exurbs, or, most dramatically, move into gated communities and apartment buildings that provide a private security force. It’s only with these enormously expensive sacrifices that we’ve eked out a slight downtick in the prevalence of crime; had we not done that, the problem would be much worse.