Chapter 36 · What Will Crypto Ultimately Become?

A Real-World Problem

Turn back to the Lab in Chapter 1 and the table you filled in.

It had two columns, bank transfers and onchain transfers. Its five rows asked who keeps the ledger, who can reverse a payment, when settlement is complete, whom to contact when something goes wrong, and whether it works on weekends and at midnight.

Your answer was correct. The number on your phone is money a bank owes you, recorded in its ledger. Thirty-five chapters later, that remains true.

But you can now see two things in that table that were invisible then.

First: it has two columns, and you now know it needs at least five or six. Cash, bank deposits, payment-platform balances, stablecoins, native assets, and onchain debt claims all answer those five questions differently, and all circulate today. What looked like a binary choice was actually a spectrum.

The second is more difficult: the table assumes something that no longer always holds.

It assumes the person filling it in is the person paying. “Whom do you contact?” assumes a “you” who can contact someone; “Does it work at midnight?” assumes a person who wants to use it then.

The last five chapters dealt with a new situation: the payer is not human. It has no identity document, cannot open a bank account, never sleeps, can pay four thousand times daily, and everything it does ultimately becomes somebody's responsibility.

The two-column table cannot contain it.

This chapter returns to the opening question with a very different answer:

What exactly is money?

This is Part A's final chapter. It introduces no new mechanisms—the previous thirty-five supplied those. It has one task: combine the six stages into a model and put that model in your hands.

Thought Experiment

Return to Chapter 1's village. We stopped in year six. Now continue.

Year seven: someone invents a new way to keep accounts. It needs neither Wang nor any trusted individual. Many copies of the ledger are distributed among villagers. A transaction counts only when enough copies record it together. Changing an entry requires convincing an overwhelming majority of copy holders—impractical in practice.

The significance is not simply greater security. For the first time, the village has a ledger independent of any particular person. Wang can leave, die, or become dishonest; the ledger remains.

Year eight: people discover that the ledger can record more than money. Land ownership, debts, and promises can all be recorded, and nobody can unilaterally change them afterward. Contracts no longer need the chief's stamp.

Year nine: agreements execute automatically. “If rice prices fall below this level, the pledged land passes to the creditor.” Nobody needs to enforce that statement: it is a ledger rule that takes effect when its condition is met. Lending no longer needs a debt collector.

Year ten: strangers across several villages jointly own a mill. Revenue distribution, expansion decisions, and control over its funds are recorded in the ledger and voted on by ownership share. This is the village's first organization without a legal representative, office, or official seal.

Year eleven: the mill starts collecting fees itself. Customers pay its account; rules distribute proceeds to shareholders. The mill is still just a set of rules. It makes no decisions.

In year twelve, something truly new happens for the first time.

Someone installs a device that observes weather, grain prices, and inventory. It decides tomorrow's operating hours and whether to buy fuel ahead of time. It decides whether to spend, rather than merely executing a fixed rule.

A new entry appears in the village ledger:

Payer: Mill
Payee: Fuel merchant
Amount: Three shells

The payer is not a person.

Villagers argue for three months. Some call it absurd: money belongs to people; how can a mill pay? Others say nothing has changed—the mill already distributed money automatically.

Both sides notice the genuine change: everything the mill previously did had been prescribed by someone. This time, in circumstances nobody anticipated, it decided to spend three shells.

A question that Chapter 1 never needed arises:

Can something nonhuman have an account? If so, who is responsible when things go wrong?

There is still no universally accepted answer. But notice: the ledger itself did not reject it. It never checks who you are, only whether the signature is valid and the balance sufficient. It treats the mill and Wang identically.

That is both what makes this possible and the source of all its risks.

Your Decision

The village assembly votes: should nonhuman entities be allowed accounts?

What Happens

The village advanced twelve years; you read thirty-six chapters. The two paths meet here.

The six stages are six parts of the same question, rather than unrelated groups of facts:

Each stage adds a layer to the preceding one. Together they answer Chapter 1's question.

Revisit the four-column framework you used to examine six forms of money:

All four columns still hold; none needs changing. Chapter 1 gave you a usable model, not merely a simplified introduction.

Three other things have changed:

First, “Who keeps the ledger?” has another answer. Previously it was an individual or institution. Now: no person keeps it; rules do, with copies distributed among strangers.

Second, the ledger records more than money. Ownership, contracts, organizational rules, outstanding debt, and market states share the ledger and can reference and combine with each other. Money is simply its earliest and simplest record.

Third, account holders need not be human. Stage six adds this layer, the one change for which Chapter 1's framework left no space.

Rewrite Chapter 1's sentence. Originally:

Forms of money differ in who keeps the ledger, who makes the promise, and when it counts.

Now:

Who keeps the ledger (possibly nobody), who makes the promise (possibly rules), when it counts (possibly determined by a program), and who holds it (possibly not a human).

Build a Model

1. The Final Model

Combine thirty-six chapters into a chain. This is all of Part A:

In one sentence:

Crypto is money, property rights, markets, organizations, and settlement native to the internet; AI adds economic participants.

“Native to the internet” carries the weight. It means four concrete things:

Participate without applying to anyone
The same rules wherever you are
Always available, with no business hours
Programs can operate it directly, without a person in between

These four properties generate both its advantages and its risks. Anyone can participate, including unwanted users; programs execute efficiently, including efficiently carrying mistakes to completion. Chapter 24 covered this other side.

2. The Six Parts

Settlement, the fifth part, is easiest to overlook and underpins everything.

Chapter 1 compared onchain transfers to cash: nobody can reverse them after confirmation. A programmatically recognizable, irreversible settlement point makes the other five parts possible.

Property depends on it; otherwise ownership remains potentially revocable. Markets depend on it; otherwise market makers cannot calculate exposure. Organizations depend on it; otherwise vote results can be overturned afterward. Software participants especially need a clear signal for their next action; they cannot merely feel that a payment is probably fine.

If Crypto must be reduced to one thing, it is settlement. The other five are applications.

3. Why the Sixth Part Is New

The first five move existing human institutions to new infrastructure. Money, property, markets, organizations, and settlement are thousands of years old; Crypto changes their medium and operation.

The sixth introduces a new category of participant instead.

The last two rows explain the risks, independently of model intelligence. A participant that never tires, does not stop itself, and performs millions of actions daily has no self-limiting errors. Multiply Chapter 32's eleven executions in forty minutes by ten thousand to see what this means.

The third row explains responsibility. Software cannot be held accountable, so responsibility returns to a person; that person must design constraints beforehand. This is the entire reason for Chapter 34.

4. Three Questions That Do Not Expire

Mechanisms change. These three questions do not. They apply to every new development, including ones that do not yet exist:

Together they reveal more about a project's position than a technical white paper. No technical knowledge is needed; you could ask them after Chapter 1.

The Analysis Toolkit's eighteen questions expand these three. If you remember only three, remember these.

5. Three Futures and How to Distinguish Them

What will Crypto become? Honestly: we do not know. We can outline possibilities and ways to distinguish them.

First: plumbing within the financial system.

Stablecoins become routine for cross-border settlement; some asset registration and transfer move onchain. Users largely do not notice, just as few care which clearing network processes a transfer.

This is the least dramatic and most likely scenario. Its signal: real payment and settlement volumes rise while discussion declines.

Second: a parallel market primarily serving speculation.

Technology advances, but most activity remains asset trading. It has real scale and participants but little connection to ordinary economic life.

This describes much of the past and could continue. Watch whether new users mainly trade rather than use products, and whether trading itself generates most protocol revenue.

Third: the machine economy's settlement layer.

Autonomous software buys services and pays other software, needing programmable, cross-border money without account opening. This is the most imaginative and uncertain scenario.

Chapter 35 supplied its signal: how many real services accept it, and why. Count actual adoption rather than specifications released.

Include another possibility: contraction.

A systemic accident, strict regulation, or simply alternative solutions to the same problems could push it into a small niche. A forecast that excludes this possibility is a sales pitch.

The key is to watch signals capable of ruling out several scenarios at once. This chapter's AI Lab practices that skill.

What It Is Called

In one sentence:

Crypto is money, property rights, markets, organizations, and settlement native to the internet; AI adds economic participants.

Chapter 1's updated answer: Money is an accounting system. That has not changed; its bookkeeper and its holder may no longer be human.

Hands-on

Money is:
Why it works:
Who keeps the ledger:
Who makes the promise:
When it counts:
Who holds it:

[Same] I already answered correctly
[Deeper] I was right, but did not know why
[Changed] My answer is different now

Money: Can I explain it myself? What am I most uncertain about?
Property rights: Same questions
Markets: Same questions
Organizations: Same questions
Settlement: Same questions
Economic participants: Same questions

Who keeps the ledger:
Who bears the risk:
What remains when subsidies stop:

The direction I most want to investigate after these thirty-six chapters:
Why: (Do not write “because it has potential”)
One concrete thing I will do in the next three months:
How I will know I have done it:

AI Lab

Predict the onchain economy ten years from now with 15 specific judgments.

Label each with exactly one category:
  [Fact] Already happening today, with a source I can provide
  [Trend] A direction observable today, but with an uncertain outcome
  [Inference] My speculation, not yet observable today

Requirements:
1. Give no specific market-size or growth figures unless you provide
   their source and definition.
2. Follow each [Inference] with “Conditions under which this fails.”
3. Include at least three failure paths: circumstances in which this
   contracts or is replaced by another solution.
4. Invent no protocol names, specification versions, or institutional positions.
5. End with a separate section identifying which of the 15 judgments,
   if wrong, would invalidate most of the others simultaneously.

Its category | My category | Why they differ

Inference | Its stated failure condition | The condition’s status today

My judgment:
Category: (Fact / Trend / Inference)
Evidence:
Failure conditions:
When I will return to check it:

Real-World Case

Change One Variable

Takeaway Questions

The twelve questions and six AI questions close this chapter. These four are Part A's exit.

Chapter Check

Part A's Final Chapter

Thirty-six chapters are complete.

You began with the number on your phone and reached whether a nonhuman entity can have an account. The distance between those questions is the entire course.

You should now be able to do three things:

First, explain why any Crypto project exists. Describe its place among the six parts, accounting structure, and risk bearers, beyond repeating features.

Second, complete independent research and specify uncertainty. Every conclusion has an evidence grade; every number a definition and date; uncertainty is explicit rather than hidden in fluent prose.

Third, define boundaries before delegating to AI. What can be delegated, what cannot, who bears mistakes, and what stops them?

Two compatible routes follow:

Complete the Capstone Project. Research a real project thoroughly. This tests usable understanding—a real delivery separates comprehension from application. Your Lab pages provide its beginning.

Build something yourself. Part B teaches reading chains, writing contracts, and constructing systems. Part A helps judge whether something deserves to exist; Part B helps you make it.

One final thought.

The industry's topic changes every six months. Your three questions do not expire: who keeps the ledger, who bears the risk, and what remains without subsidies. When an unfamiliar development appears, ask these first.

Most things fail them. The few that pass deserve serious attention.