The new set of tools the last chapter pointed to has a name, and the name has become a problem in itself. Web3, crypto, blockchain: the words now arrive pre-loaded - with fortunes made and lost overnight, with scams and celebrity coins and the general atmosphere of a casino. It is fair to say that much of what happens under these words deserves the skepticism. But there is a difference between what people do with a tool and what the tool is, and if we want to think clearly we have to make that separation on purpose. Strip away the speculation, the price charts, the hype - and look only at what the underlying machinery actually does. What is left is more serious than the casino around it, and it speaks directly to the problems of the first two chapters.
Begin with the core of it. A blockchain is, in plain terms, a shared record kept at once across many independent computers that agree on its contents, where entries once written are practically impossible to alter or erase, and where no single party administers the whole. That last clause is the important one. The record is not held in a company's database, subject to a company's discretion. It is held in common, and its rules are enforced by the network rather than by an owner.
Set that against the first chapter. The hierarchies described there depend, every one of them, on being a central point through which things must pass - a ledger someone controls, a registry someone keeps, a switch someone can throw. A system with no such central point is a system with far less of the leverage those hierarchies run on. And because the rules live in open code that anyone can read and audit, they are, at least in principle, the opposite of the closed back room: not "trust us," but "check for yourself." For a world whose institutions have taught people, reasonably, to assume the worst about what they cannot see, a machinery that can be inspected instead of trusted is not a small thing.
The second thing these tools do lands on the individual, which is where the second chapter left us.
They let a person hold what is theirs directly. In the ordinary financial system, your money is a number in someone else's ledger; the institution holding it can freeze it, delay it, or be compelled to. With self-custody, a person holds their own keys and their own assets, answerable to no intermediary's permission or solvency. In the same way, these networks are permissionless: you do not need anyone's approval to open an account, to transact, to build on top of them. There is no gatekeeper deciding who is allowed in, because there is no gate. And they are censorship-resistant: a valid transaction cannot easily be singled out and blocked, because no single party holds the position from which to block it.
Put plainly, these are tools that turn the individual from a supplicant into a holder. The person who was, in the last chapter, a subject - dependent, permissioned, revocable - is handed a form of direct agency that does not ask permission and cannot be quietly switched off. That is a real answer to a real part of the diagnosis, and it should be named as such.
The third thing is the most interesting, because it speaks to the exact specification the last chapter ended on: how do people act together at scale without building a new machine over their own heads?
These tools make possible a form of collective action with no boss. Groups can pool resources, put proposals to their members, record the decisions in the open, and have the agreed rules carried out as written - all without a central authority that sits above the group and can override it. The organization is not a building with an executive inside it; it is a set of rules the members hold in common and can watch in operation. Coordination, in other words, without a central point of control - the same property that mattered at the level of the ledger, now applied to the level of the group.
This is genuinely new. For most of history, acting together at any size meant erecting a hierarchy to hold the coordination - and the hierarchy, once built, tended to become the very thing that turned participants back into subjects. Here is at least the raw material for something else: assembly that does not, by construction, require a master. It is the first candidate answer we have met to the chapter-two problem, and it is a serious one.
I want to state the verdict without hedging, because the argument of this book depends on my being fair here before I am critical. These are not toys, and they are not merely instruments for speculation. Underneath the noise, they attack precisely the things the first two chapters identified as the disease: the central points of control, the gatekeepers, the dependence of the individual on institutions that need not answer to them. Anyone serious about the concentration of power should take them seriously. I do - and the project this book describes is built with them, not against them.
And yet.
Notice the shape of everything just praised. Each strength is a removal. No central administrator. No gatekeeper. No permission. No single party who can block, freeze, or override. These tools are, at their core, extraordinarily good at taking control away. That is exactly what a world of overbearing hierarchies seems to need, and it is why the enthusiasm is not foolish.
But taking control away is not the same as putting something in its place. Removing the master does not, by itself, create accountability among the people left in the room. Removing the gate does not, by itself, let a durable and legitimate community form inside. Strip a space of every point of control and you are left with freedom - and freedom is the right word, and a good thing. But freedom in a space with no light and no floor beneath it is not merely liberating. A room with the locks removed and the lights off is open to everyone, which means it is also open to anyone, and what fills a space like that is not always what its designers hoped for.
That is the other edge of the tool, and it is the subject of the next chapter. Not a retraction of anything said here - the tools are as good as I have claimed - but the recognition that a pure architecture of removal has a characteristic failure, one the movement's own enthusiasm tends to look past. And the clearest place to see that failure is not in theory. It is in a place that took these same properties further than Web3 ever has, and shows plainly what they produce when nothing else is added.