wallet/docs/white_paper_YarvinAppendix.md

---
title: >-
    Bitzion: how Bitcoin becomes a state
...
It probably won’t happen. It probably should.

Statelike nonstates fascinate all political engineers. Can a nonstate
become a state? How are states born? What paths, if it wants to live,
should a baby state follow?

History is full of examples. But most of them are too ancient for most
of us to parse. The vulgar and legible present is frustratingly short of
high-quality newborn states.

One path to statehood is military. The CJNG or Jalisco New Generation
Cartel, a Mexican narco gang, recently posted a video of an impressive
military parade, with hundreds of uniformed soldiers and tens of armored
vehicles—almost ISIS tier. ISIS in its day, as Gladstone said of the
Confederacy, “had made an army, and made a nation.”

But having made these things—a state must also keep them. Vae victis!
In the end, sovereignty is empirical. Cartels, ISIS, the Confederacy and
Rhodesia aren’t super encouraging examples.

Yet after this list of almost-states comes very solid and palpable
nations, like Israel, Singapore and our own dear United States.
Empirically, states do get born.

So, empirically, is Bitcoin a state? Or almost a state? Can it become a
state? Or at least, a virtual state?

Bitcoin has not yet made a nation. It has certainly made a treasury. Its
path is not at all military—except if math is a weapon. But isn’t math
a weapon? Surely math has as good a chance as the CJNG, and you wouldn’t
want to mess with the CJNG. “Pura gente del Señor Mencho!” In fact,
Bitcoin and the CJNG even have similar market caps...

Three attributes of sovereignty

When we examine the early life of successful states, we see that newborn
regimes which live to become stable adults tend to focus on three
attributes: independence, legitimacy, and coherence.

Independence means the regime does not depend on the will of any outside
power. Legitimacy means that the regime considers itself authentic and
official, and conducts itself with the dignity of a state. Coherence
means the whole state can act effectively as a single agent, with a
united regime calling on the full resources of all its citizens.

In the ways that Bitcoin works well, it has all three. Abstractly, the
ledger itself is independent, legitimate and cohesive. But as a new
state rises, reality tests it more and more stringently. Some respond to
hard tests by becoming stronger. Others crumble.

Bitcoin today has cracks in all three attributes of statehood. These
cracks may or may not need to be filled. Filling them may or may not
result in any kind of success. We can still describe how they should be
filled—as a theoretical lesson in political engineering.

We can also describe how Bitcoin, with these vertical cracks filled, can
also expand horizontally. It will both increase its effectiveness, and
expand its scope. We’ll project this growth outward until Bitcoin turns
into Bitzion—a purely hypothetical utopia, which needs no fiddling
with definitions to be a clear and obvious digital state.

Historically, Bitcoin has been extremely resistant to innovation in
governance. And, as we’ll see, the principles of effective digital
governance are extremely counterintuitive. This makes it unlikely that
anything like Bitzion will actually happen. But since it could happen,
it makes the best possible kind of thought-experiment.

Bitcoin: from algorithm to state

Again, while Bitcoin has made neither an army nor a nation, it has made
a treasury. This treasury is its market cap, which indeed resembles the
balance sheet of a minor state—without any of the assets. But having
made a treasury, it must keep it.

One way to price Bitcoin is as a call option on monetary sovereignty. If
Bitcoin becomes a state, or merges with a state—becoming either a
standard hard currency, or a national or global hard currency, de facto
or de jure—we can expect the option’s price to considerably increase.
If this outcome grows less likely, the option’s price will fall.

Are “becomes a state” and “becomes a standard” really equivalent?
Certainly not without an argument—and that argument is the crux of
this essay.

The difference between Bitcoin as a standard, and Bitcoin as a state, is
that a state has a government. The argument: to prevail as a monetary
standard, Bitcoin must up its game. Its most effective way to up its
game is to form its own government. But once it forms a government, it
is a state—contradicting its libertarian, decentralized ideology.

Interestingly, the 13 states faced a similar question in the mid-1780s.
The birth of a centralized national government contradicted much of the
ideology of the Revolution. Yet the Congress of the Confederation was
such a shitshow—almost an 18th-century UN—that the
pre-Constitutional United States has been almost written out of history.

The case for Bitcoin as a state is that a monetary standard with a
government can work much better than one without one. Bitcoin already
has a government; but it is (mostly) algorithmic. A human government can
work much better than an algorithmic one.

All Bitcoiners agree that not having a government is an advantage, since
a government is inherently a central point of failure. While this is the
truth, it is not the whole truth.

Having a central decision point is also an advantage. As hackers, we
hate choosing between two such advantages. We would like them both. And
if we cannot have all of both, perhaps we can at least get most of both.
Such is the good old art of the tradeoff.

Logical case for a Bitcoin government

Why does Bitcoin need a (human) government?

Bitcoin at present is a state without a government. If there is one
thing political scientists know, it is that there is no such thing as a
state without a government. When a state tries to have no government, it
always has something.

A blockchain is statelike because independence is the whole point of a
blockchain. Bitcoin or any other blockchain has two layers of
governance: serialization, which is inherently algorithmic (execution of
the ledger protocol); and oversight, which is inherently human
(management of the protocol and its reference implementation).

Government is human; it means oversight. Bitcoin needs a human central
government because centralized governments are more effective than
decentralized governments. An effective government can be trusted to
choose the best serialization algorithm; so algorithmic serialization is
just an aspect of human oversight.

Bitcoin today has serious oversight problems. No one even agrees on who
has the responsibility of overseeing the blockchain—miners, nodes, or
developers? This is why Bitcoin is a state without a government.

The only correct answer is hodlers—who are the only group not now
represented. They could be represented only by constructing a central
government that represents them. This is just what they should
do—though it would change the whole story of Bitcoin.

But since Bitcoin’s independence today is secured by decentralization,
any central government would need a different security mechanism to
secure its independence. One such mechanism is pseudonymity.

Pseudonymity has many technical limitations and is a broadly inferior
substitute for decentralization. Engineers, in pursuit of the perfect
tradeoff, must often use limited and inferior substitutes.

Intuitive case for a Bitcoin government

Ultimately, what is Bitcoin? It is a set of human beings with private
keys, who all use a technical tool that lets them automatically agree on
digital facts. Bitcoin is the people, their keys, and the ledger—the
public record of how much money every key owns.

Bitcoin’s algorithm and infrastructure are a cool tool that these people
have used to act coherently for their mutual benefit—without uniting
or cooperating in any other way. Nakamoto consensus is a rope of sand
that can actually pull a load. The hodlers, a completely atomized
non-society, nonetheless act coherently when they save in coins.

Collective coherent action for mutual benefit is a tool of essentially
unlimited power. Algorithmic, decentralized governance has unlocked a
tiny slice of this power. The normal way in which sets of human beings
undertake collective coherent action is through central organization.
This normal government can unlock the rest of the pie.

Bitcoin, by producing governance without a government, is the exception
that proves the rule. What this exception tells us is that whatever
Bitcoin has achieved so far, these achievements are a small fraction of
what it could achieve with a government.

While this magic rope of sand is cool and all—while Bitcoin could not
have been born without it—what about switching to, like, a regular
nylon rope?

Must this be done? Why must it be done? Can it be done? How would it be
done? If you have a moment or two, dear reader, we’ll answer these
questions and more.

Bitcoin’s oversight is a mess

Bitcoin has a reputation for terrible oversight. This reputation is not
the fault of anyone involved. It is Satoshi’s fault, for leaving
oversight to anarchy. It is not even clear which group holds the power
of oversight: nodes, miners, devs, hodlers or users. (Satoshi was a
genius; but not a god.) Fact: anarchy can make anyone behave badly.

The only proper interpretation of Bitcoin oversight is that Bitcoin’s
principals are its hodlers. Bitcoin is the ledger. The blockchain is an
agent that maintains that ledger. The Bitcoin infrastructure must be
operated in the exclusive service of the ledger, just as any agent must
act in the exclusive service of its principals. Had Satoshi made this
clear, or even built a mechanism to enforce it, much suck might have
been averted.

Anarchy is not the end of the world. But anarchy still sucks.
Objectively minor meta-decisions, like setting the block size, turn into
Icelandic sagas. Larger changes are inconceivable. Bitcoin, the OG of
blockchains, is the couch potato of blockchains. Everyone knows this, so
there is no reason to dwell on it.

Bitcoin’s serialization isn’t perfect, either

Bitcoin’s serialization—maintaining the integrity of the ledger—has
a perfect record. No power has ever infringed the independence of
Bitcoin. Yet hash proof-of-work is not perfect. It is not perfectly
secure; and it is extremely expensive.

Today, 65% of the world’s mining capacity is in China. So it would be
straightforward for the Chinese government to squeeze Bitcoin. The CCP
has not felt the need to try—not because Bitcoin is too strong to
attack; because it is not important enough to attack. This is not
independence but the illusion of independence. If Bitcoin is indeed an
call option on sovereignty, these kinds of illusions are dangerous.

The cost of mining is the price of independence. Currently this price is
huge: perhaps 5% of market cap per year. Imagine gold had a high vapor
pressure—every year, 5% of every Krugerrand boiled off into the
atmosphere. In this alternate universe, no one would even think of using
gold as money.

Bitcoin hodlers do not think they are paying this immense wealth tax.
They are. And it seriously hampers the competitiveness of Bitcoin as a
currency. Let’s look closely at how this works—these details matter.

The price of Nakamoto consensus

Mining revenue comes from fees and block rewards. Block rewards are
straight-up monetary dilution. Proper accounting must measure ownership
of any equity or cryptocurrency not as the number of units, but as the
fraction of units outstanding. So creating 1% more units is exactly the
same as a 1% wealth tax across hodlers—though easier to implement.

Block-reward dilution is bounded, and fees are not dilutive. But fees
have the same direct impact as rewards: since they need to pay for
electricity, they are sold for fiat. Miners are generally not hodlers.
So \$5B of mining means \$5B of selling. Obviously, the seller is the
natural enemy of the hodler.

As a currency—a stable bubble—Bitcoin acts like a battery or a
pressure tank. Its market cap is the pressure in the tank, or the energy
in the battery. To counteract \$5B in annual outflow and keep the price
flat, Bitcoin needs \$5B of annual inflow.

This mining leakage gets heavier as the price rises—a vicious
stabilization loop which tends to cap the price. This is how the cost of
governance by Nakamoto consensus costs the hodlers money—and tends to
make hyperbitcoinization impossible.

Imagine Bitcoin’s price if it had no mining leakage—meaning zero
automatic outflow. Everyone selling would be a defecting hodler. There
has never been a large number of these, even in the worst crashes.

Algorithmic governance was essential in Bitcoin’s childhood. It created
a system that was owned by no one. From its genesis block it had this
essential quality of a state. Nakamoto consensus made Bitcoin
independent from day one. No other design could have achieved this
result.

But statehood has its own ruthless laws. No startup can cling to its
teddy bear. All must change as they grow. Some of those changes will
even be in engineering. Today, the cost of Nakamoto consensus is keeping
Bitcoin from advancing as a monetary standard; and its unbalanced
decentralization leaves it under China’s thumb.

Bitcoin imperialism

By all except Bitcoin’s proof-of-work purists, it is generally admitted
that proof-of-stake is the path forward in both layers of
governance—serialization and oversight.

Many alternate chains already use POS—an unfortunate acronym—as well
as smarter agreement protocols, chain sharding, and all kinds of other
fun stuff. (In fact, it would be hard to boot up a new proof-of-work
chain today—it would be attacked as soon as it was worth attacking.)
Why not just switch to a newer-generation blockchain platform?

Alas, none of these alternate chains can solve the problems with
Bitcoin. Instead, Bitcoin will have to defeat them and erase them
utterly from the earth.

Most Bitcoiners believe this victory is in some sense inevitable, and
will happen on its own. Is it happening on its own? Or is anyone
thinking about how to make it happen? This is the mindset of Augustulus,
not Augustus—a poor start for any empire.

For it does need to be made to happen. It is not just a matter of
waiting for altcoins to go away. Rather, Bitcoin needs to actively clean
them up.

This uber-hardcore flavor of Bitcoin maximalism—call it Bitcoin
imperialism—would probably shock even most Bitcoin maximalists. But it
is an inherent consequence of the right theory of why Bitcoin is worth
anything at all.

Parallel ledgers considered harmful

An alternate chain is a parallel ledger. All ledgers must be unified.
Any partition of monetary energy makes not only each side of the
partition, but both sides put together, a worse bet than the unified
ledger. And there is no stable equilibrium between the sides of the
partition.

In any real economy, at least one asset must be overvalued. That asset
is money. But given one money that is both standard and perfect, the
only stable equilibrium is one in which there are no other bubbles, and
all other assets are priced by yield in that standard. Any such asset
which has no use has no value, and will sell for no price.

There is one mathematical revolution in money. Forks and altcoins are
just bad for it. The revolution in money will be complete when there is
one standard digital money. This means one standard ledger—the
“Highlander principle.”

This standard ledger has to begin with Bitcoin. Bitcoin is the gorilla.
In any world where Bitcoin ends up worthless, lesser cryptos end up
worthless too. So there are two ways to resolve any parallel ledger:
merge it with Bitcoin, or send it to zero.

Generally, any alternate chain worthy of any notice at all has some
technical advance, purported or real, over Bitcoin. This may be privacy,
computation, finalization time, mining leakage, Byzantine security, or
some other useful area of improvement.

But the real competition is monetary competition between ledgers. This,
it is hard to imagine Bitcoin losing. Muscle matters more than mind. The
chimpanzee can ask for a banana in sign language. The gorilla is three
times as big as the chimpanzee.

The traditional Silicon Valley approach to this conundrum is for the
gorilla to just kill the chimpanzee and steal his banana. Sometimes, the
chimp can jump the gorilla from behind, and get him with a rock.

Neither outcome seems quite right. A cryptosystem that aims at the
legitimacy of a state cannot behave like any kind of ape. We’ll return
to this issue with an actual plan. However: there can be only one.

Proof-of-stake and political engineering

Altcoins delenda est. But while their parallel ledgers must vanish with
the Oscans and the Samnites, as their genuine engineering innovations
are shamelessly looted like Houston lifting the Nazi space program, many
have also innovated in governance. Scholarship asks us to start by
noticing their work; and we will start by following it.

While proof of work is one thing, proof of stake is a huge family of
things. This family is united by one general principle. The general
principle of proof-of-stake is that the ledger belongs to its
owners—who must therefore be charged with governing it.

That principle is sound. Bitzion uses it. But it is as general as the
principle that the nation must be governed by its citizens—and as
vague. Even if everyone agrees on democracy, there are infinitely many
possible democratic constitutions.

Proof-of-stake is not democratic, of course, because power is by wallet,
not by head. The ledger belongs to its owners, in proportion to their
ownership. The ledger is managed for the benefit of its owners, in
proportion to their ownership of it.

The proper term for voting by wallet is plutocracy. That this is a
pejorative makes it all the more delightful. But a plutocratic republic
is still a republic—and faces the same general challenges in political
engineering that vex any democratic republic. Here we will say democracy
when we mean plutocracy. Since head-counting is impossible, there is no
ambiguity; and euphemisms are nice.

Most proof-of-stake altcoin designers are not trying to innovate in
political theory. But their political engineering is still
awful—because it comes from the awful fashions of an awful period, in
which they happen to live.

Why would they realize that following the fashionable ideas of their era
would produce awful results? If everyone is using the same bad ideas,
everyone’s results will be awful, and no one will notice. They will just
think the problem is very hard—which it is.

Let’s look at how to take this principle, proof-of-stake, in an unusual
direction. Instead of building ideal governance structures, let’s
reengineer existing governance structures.

These designs are not fashionable—since they predate the 20th century.
No one would invent them now. But everyone still inhabits them. Isn’t
that weird?

If our fashionable ideals are realistic, this plan will just mean
inheriting ancient cruft. If today’s ideals are not realistic, it may
produce dramatically improved governance.

From Bitcoin to Bitzion

To turn Bitcoin, the ledger and blockchain, into Bitzion, the digital
state, we adopt the plutocratic principle of allocating authority and
responsibility by stake. We declare that Bitcoin is the ledger, not the
protocol or the infrastructure. We declare that the ledger is owned by
the hodlers; it should be governed by the hodlers, for the hodlers; and
we, the hodlers, will cultivate it into a virtual state that will
astonish the world.

Although it uses proof-of-stake, Bitzion makes three large theoretical
adjustments to the viewpoint of most classic altcoin-style POS
governance designers. These are not technical changes; but they lead
toward completely different technical challenges.

First, most designers focus on ledger governance (algorithmic
serialization); we’ll focus on systems governance (human oversight).

Second, most designers consider direct democracy the ideal form of
democracy. Actually, indirect democracy creates much more effective
governance structures.

Third, most designers think decentralization is the only tool for secure
independence. But there is another such device: pseudonymity.

Serialization, oversight, and trust

Serialization means recording a fair, unified, reliable and independent
ordered list of transactions. Serialization is the job of every
blockchain. Centralized, trusted serialization is easy. Decentralized,
untrusted serialization is hard. So serialization is what most
cryptosystem architects think of, when they think of governance.

Of course, humans cannot be in the serialization loop. Serialization is
inherently algorithmic, if just because it has to be fast, secure and
scalable, as well as fair (it does not censor or reorder transactions)
and independent (no outside force can influence it).

Here is a small philosophical crack in the thinking of most
cryptointellectuals. Why is Bitcoin like a state? Because it is fair,
reliable, and independent. Why is it fair, reliable, and independent?
Because it is decentralized. We all agree on this, but—

It is easy to overlook the small but important fact that what users
value in Bitcoin is not its decentralization. What users value is a
consequence of its decentralization: that Bitcoin is fair, reliable and
independent. More broadly: that it is legitimate, effective, and
sovereign. Still more broadly: that it is good at being like a state.

The Bitzion thesis is that Bitcoin has ways to become much better at
being like a state. If Bitcoin is an option on statehood, this should
make its price much higher. Any tactical path that produces this result
must be at least worth considering.

It is essential that we understand this difference between cause and
consequence. We trust that Bitcoin is fair, reliable, and independent,
because math proves that its decentralized protocol makes it so.

Or so the nerds tell us. Most of us cannot do the math—we probably
don’t even know what an elliptic curve is. Byzantine what? But we trust
the nerds. Trust is beautiful.

Suppose we trust that Bitcoin is fair, reliable, and independent,
because it is run by the god Apollo. Apollo keeps the ledger on a big
SCSI RAID rack in his basement, with a nest of ribbon cables running
into a Compaq 386 PC-XT. It’s as fast as Nakamoto consensus, and even as
reliable—because he’s Apollo, and Apollo is perfect.

Or so the priests of Apollo tell us, and we trust the priests. Trust is
beautiful. The only relevant statement in both cases is: “we trust that
Bitcoin is fair, reliable, and independent.” The cause of trust, whether
decentralization or Apollo, makes no difference. What counts is the
trust itself. (It’s better if the trust is rational, though.)

Besides serialization, which is governance in the proper blockchain
sense, all chains have parameters; all chains have protocols; some have
full operating systems. The power to update these rules, values, and
systems is oversight.

Oversight is an inherently human function—though the identity and
procedure of those humans may depend on the state of the ledger. Since
all blockchains require some oversight, there is no choice but to trust
the oversight nerds. Trust is beautiful.

As previously mentioned, oversight in Bitcoin has historically been a
shitshow. In POS altcoins, extending voting by stake from serialization
to oversight is a natural idea. But that makes oversight a natural
afterthought.

Bitzion attacks the same problem from the other end. It reasons:
effective oversight can choose the best serialization algorithm. Get
oversight right, and oversight will get serialization right.

Again, oversight is just government. The solution to the serialization
problem is: form an effective government, and trust it to develop and
manage an efficient serialization network resilient to attacks by global
adversaries.

Animal control is a natural function of physical government.
Serialization is a natural function of digital government. Both powers
can be extremely dangerous if misused.

Malicious serialization cannot forge transactions, but it can censor and
revoke them. Malicious animal control can dart and capture humans,
bundle them into white vans, do some processing in an underground lab,
and peddle the organs on the dark web. Don’t a lot of people just
disappear? What do you think is happening to them?

When citizens of a physical government think about how to keep bears and
leopards off the street, without opening the door to rogue dogcatchers
who secretly moonlight as human organleggers, they tend to focus not on
this specific state function, but on delegating their consent and
loyalty to a whole government which is stable and sane. Citizens of a
virtual government also trust that their stable and sane government will
operate a secure and efficient serialization mechanism. Again, trust is
beautiful.

Since Apollo isn’t returning our calls, any serialization network will
almost certainly be decentralized. The government that creates it, since
it must be highly effective, is likely to be... centralized. Obviously,
to any cryptointellectual, this is a huge red flag.

By default, any central organization is a big step away from
sovereignty. Independence is one of the three attributes of sovereignty.
A decentralized non-organization is very hard to pressure. By default, a
centralized organization is very easy to pressure—

By default. To apply the lesson we just learned, decentralization is not
a design goal of a cryptostate. Decentralization is a mechanism. The
purpose of this mechanism is independence. The purpose of independence
is sovereignty—or at least, monetary sovereignty. So as engineers, we
are free to use other mechanisms of independence.

But independence, in Steve Jobs’s words, is a feature—not a product.
Let’s design our new regime, then figure out how to make it independent.

Indirect democracy: magic in plain sight

The main assumption that makes proof-of-stake governance ineffective is
the almost wholly false assumption that direct democracy is better than
indirect democracy.

This is a weird belief in a civilization whose realization of democracy
is, and has always been, indirect. Democracies are rare in history. But
direct democracies are extremely rare. Nor has the judgment of the past
given them great reviews. When Athens was a direct democracy, it was the
assembly who voted to execute Socrates.

But direct democracy is more consistent with 20th-century ideology. We
feel, quite reasonably, that indirect democracy is one of our many
quaint legal anachronisms.

Obviously not everyone had Internet in the 1780s. They had to use
dialup, or horse-drawn ballot wagons, or something. Even now, attention
is in limited supply—but, at least in theory, every citizen should
have the right to vote directly on every decision. Also, politics sucks
and politicians suck. That’s the normal perspective.

We love democracy and hate politics. When we realize that our new
cyberstate can have democracy without politics or politicians, since it
can count zillions of ballots in zilliseconds, this seems like an easy
win. We don’t even think of doing it the old way, even though
everything in the real world works the old way—indirect democracy.

Actually, direct democracy sucks. Virtual states will be weak and
passive at best, chaotic and ineffective at worst, until they somehow
reinvent indirect democracy.

A jury is not a government

All systems of direct democracy, including all proof-of-stake designs
that I know of, use their stakeholders as a jury. A jury is a group of
people who vote to ratify or reject decisions framed and proposed by
some other process.

A legislature is a jury in which the members themselves have the right
of proposal. All actual legislatures either delegate this power to
external structures, or evolve internal structures (a cabinet in the UK,
committees in the US) which are hierarchical.

Direct democracy just expands the jury to the whole electorate. While
nowhere in the world, except perhaps some Swiss cantons, even pretends
to be governed by direct democracy, many jurisdictions have adopted the
referendum as a decorative bauble.

Digital governance can trivially borrow the referendum, and count itself
democratic. A proof-of-stake blockchain can use a jury, full or
sortitioned, for serialization and/or oversight. Even Bitcoin oversight
only differs from POS in that the jury is the nodes.

Actually, juries suck. The jury is the most passive and reactive form of
government that can be devised. It is so reactive that it is not even a
complete government—it lacks any real executive arm.

Any vacuum in sovereignty will fill itself. A sovereign jury evolves its
own informal leadership. When Parliament replaced the Crown as the locus
of English sovereignty, to replace King and Privy Council, it evolved a
Prime Minister and a Cabinet. This was not anyone’s plan, or anyone’s
theory—it happened only because it had to happen.

Luckily the delegation from MPs to this pseudo-king, the PM, was
well-structured. Normally, the outcome of the sovereign-jury design is
an illegitimate and informal leadership agency, plus a jury which at
best is a rubber stamp and at worst a zoo.

What the jury design is best at: making large numbers of people feel
important. As members of the jury, everyone can feel in charge of
everything. This potent narcissistic energy is the principal marketing
advantage of direct democracy—and what better moment for it to shine,
than our historic golden age of narcissism?

If a machine was designed to minimize generated power, while maximizing
perceived power, it couldn’t be better designed than a gigantic jury.
Let’s illustrate the difference with a design which minimizes perceived
power, and maximizes generated power.

The political amplifier

Indirect democracy is not an anachronism. The reason that indirect
democracy is more common than direct democracy is that delegation is
amplification. Indirect democracy generates concentrated, coherent
power. Direct democracy does not.

Delegation is amplification because delegation concentrates and focuses
trust—a laser for trust. A weak, coherent laser is stronger than a
strong, incoherent flashlight. You can look at a 40-watt bulb. A
40-milliwatt laser will rip a new asshole in your eye.

A mob, in which every participant retains freedom of action at all
times, is much less effective than a much smaller army, whose soldiers
delegate their freedom of action upward to one commander. The difference
between a mob and an army is so strong and clear that it suggests the
analogy of turbulent versus laminar flow.

In indirect democracy, citizens delegate power to politicians. Citizens
cannot act directly, only by voting. There are fewer politicians than
citizens; so elections must concentrate power. But voters can use power
only by deciding whom to give it to.

Even in our indirect democracies, though their rational choice is
between politicians, voters still think more about issues. They seem to
feel they can somehow act through the politicians they elect—though
there is little evidence that this works. On paper, their democracy is
indirect. In their hearts, it is direct.

So while they love democracy, they hate politics—or to be exact, they
hate their enemies’ politicians, while merely disliking their own. Can
we blame blockchain designers for starting from an ideal, rather than
reverse-engineering this mess?

In direct democracy, where there is no delegation, every citizen can act
directly on every issue. There is no need for an illusion of
indirection. You actually matter. Your voice is heard. And there are no
politicians.

But you are collectively weak—because you, the people, are a mob and
not an army—a bulb, not a laser. If an electorate as a whole does not
matter—just because it is not strong enough to keep its dominance, or
even relevance—no one’s voice matters at all.

So the narcissistic desire for power, which is the desire to feel good
and powerful, and the ethical desire for power, which is the desire to
see power do good, are generally opposites. Almost always, the strongest
thing to do with power is not to use it, but to give it
away—concentrating it in the hands of someone who can use it more
effectively.

So when we measure the power of an indirect democracy, we measure the
amount of power that citizens delegate to politicians. This is the
delegation efficiency of the election. If this is 1, the politicians
have the unconditional loyalty of everyone who voted. If it is 0, voters
voted for random, ironic, frivolous reasons, and didn’t actually care at
all. And again, delegation is amplification.

Here’s an example of how constitutional designers can adjust this knob:
the length of a politician’s term. In the electronic world, terms could
last any length; they could as easily be a week. When we shorten terms,
does amplification decrease, or increase?

It decreases, of course. When we elect a President every week, we the
citizens are saying to the election’s winner: we trust you to serve,
but just this week. We have our eyes on you! Right now, we definitely
think you’re hot stuff! See you next Sunday!

Obviously, if you have to run for President again next week, and you
have no way to be sure anyone won’t change their mind about you, you
are on a pretty tight leash. The shorter your chain, the weaker you
are—simply because the voters who voted for you made a much smaller
commitment.

So shorter elected terms mean weaker delegation, which mean weaker
amplification. Let’s go to the other extreme. We’ll elect a President
not for four years, but forty. Or why not\... life?

If we mean our votes sincerely, when electing a President for life, we
are making a permanent political promise—half election, half marriage.
The candidate for whom we are voting can expect our unconditional and
lifelong loyalty and obedience. (Of course, our Supreme Court, for whom
no one ever voted at all, is in exactly this position.)

Surely any President-for-life will be much, much more powerful than any
President-for-a-week. So the delegation is stronger. So this design
gives the voters themselves more power—since the more power their
delegates have, the more the voters have.

By going from an election every four years to one every forty years, we
concentrate the power of ten elections into one. At this point perhaps
the awful irony is almost clear. But let’s go even farther. How would a
constitution generate maximum amplification? How would we design the
most powerful possible election?

This election would elect not many politicians, but just one. No
organization is so big it needs multiple leaders—in fact, the bigger
it gets, the more important unity becomes. So the President would have
unconditional control of the whole government. Crazy! Also, optimal for
that attribute of sovereignty we were talking about—coherence.

And there would be no term at all, not even life. The President would
designate a successor. This would be the most powerful election
ever—and, of course, the last. Until the next revolution, of course.

So... we’ve just shown that absolute monarchy is the strongest form of
democracy. An interesting result, n’est ce pas? Wily intellectuals have
all seen proofs that 2+2 is actually 5. They may amuse themselves by
checking the above logic once or twice.

It is also worth noting that the constitution this election would create
is merely that of the Roman Empire under the Antonine emperors—which
Gibbon, 1500 years later, felt was the best government humanity had ever
experienced.

Amplification versus sensitivity

This is not a suggestion that either Bitcoin or America should adopt the
constitution of the Antonines. Either could certainly do worse; both can
probably do better. But we are just illustrating a theoretical tradeoff
that often goes unobserved.

This is the tradeoff, in any republic, between sensitivity and
amplification. Sensitivity is how accountable the leaders are downward
to the will of the voters. Amplification is how much energy the voters
delegate upward to the leaders.

Our one-time election maximized amplification and minimize sensitivity.
Our weekly elections maximized sensitivity and minimized amplification.

When designers opt for high sensitivity, their decision suggests a
powerful fear by themselves or their customers of a failure mode in
which agents betray principals. Government is a dangerous agent and must
be tightly leashed.

But when principals micromanage their agents, the leash is so tight that
it chokes—and the agent cannot serve at all. You can walk a dog on a
leash. You can’t hunt with a dog on a leash. The optimal leash is the
leash that does not even exist, except in the unlikely case where it is
actually necessary—and in that case, the leash is inexorable.

We are the hodlers. Government is our dog—but a dog that must hunt.
Fortunately, we have the technology to let a dog run free and wild as a
wolf, until its owners blow a magic whistle. This recall-only republic
will be low in sensitivity, high in amplification; but its sensitivity
is not zero. It is not (quite) the constitution of the Antonines.

Pseudonymity versus decentralization

Bitzion needs to retain Bitcoin’s independence and would ideally
enhance it. How can a centralized organization achieve this?

Simple: by replacing open decentralization with closed pseudonymity.
Decentralized open networks are hard to kill because they are built out
of many small fish, each hard to catch. Closed pseudonymous networks can
be built out of a few big fish, each almost impossible to catch.
Satoshi’s identity is still a secret—so it can be done.

Again: decentralization is the means; independence is the end.
Pseudonymity is trickier and more complex than decentralization. A
gasoline engine is trickier and more complex than a steam engine. It has
many more ways to go wrong. Sometimes the simplest answer ends up being
the right answer. But you don’t drive a steam car.

Independence is inherently a security problem. In all security problems,
we need to begin with a threat model.

The threat is a global adversary—a government or quasigovernmental
force—which wishes to involve its will in our governance—to pressure
us. Independence is boolean. Unless this pressure has absolutely no
effect, our independence is compromised. History suggests that
independence, once compromised, is quickly rent asunder.

The power available to this adversary is not a constant. It will tend to
increase over time. It also increases as a function of the nefariousness
of its prey. Therefore, one core security measure is
legitimacy—maximizing compliance, both legal and social. Any future
state demonstrates heaven’s mandate by complying both with its own
ethical standards of tomorrow, and the bizarre whims of today’s armed
clowns.

Suppose our closed pseudonymous system is a small committee—with
between five and fifty members. Groups this small, if well-organized,
can act with perfect cohesion. Ownership of a pseudonym is possession of
a key, ideally sealed in a physical object—like one of Tolkien’s magic
rings. Strong pseudonyms are synthetic—random syllable or word
strings, never nicknames.

Can a global adversary find any of the keybearers—assuming all their
opsec is strong, and so is their compliance? Its best bet is to trace
their patterns of communication. So their best bet is to communicate as
little as possible. It is quite hard to trace occasional signed
statements dumped onto the Internet.

And when keys are literally passed from hand to hand, tracing custody
chains is hard, even for a global adversary. Even if the first link in
the chain is a known individual, one link that refuses to break, cannot
be found, has randomized the transfer, etc, sends the adversary back to
looking for the physical origin of an encrypted communique.

Can pseudonymity deliver a more secure level of independence than
decentralization? With a majority of mining capacity in China,
decentralization is not very secure at all. What if all these keybearers
were in China? Could they still communicate securely? It would be
difficult—but possibly doable.

In today’s West it is certainly doable. For economic reasons, China is a
great place to mine Bitcoin—Satoshi never realized that the easiest
way for a state to capture Bitcoin would just be to suck in the most
miners with the cheapest electricity. China is not a great place to be
part of any kind of secret organization, however legit. So probably none
of the keybearers would be there.

What if the Western Internet became as well-controlled as China’s?
Politically, this is not utterly implausible. Technically, it would
involve Western governments being as competent as China’s. This couldn’t
happen right away. There would be time to evade. And it is impossible to
devise a plan so perfect that it always wins without a struggle.

Bitzion in a nutshell

Bitzion is a design for a human government for Bitcoin. The goal of the
design is to be independent, legitimate, and coherent—as befits a
virtual state.

(“Zion” can be interpreted in many senses—Biblical, Israeli, Mormon,
Wachowskian or, of course, Rastafarian. The analogy to the LDS may be
especially apropos.)

The fundamental principle of Bitzion is that Bitcoin is the ledger.
Therefore, Bitcoin is the property of its hodlers. It must be governed
with their exclusive consent, and in their exclusive interest. No one
else gets a vote—“no god or government.”

Miners, nodes, developers, and even algorithms are infrastructure that
serves the needs of the hodlers. Their opinions will always be
interesting, relevant and valuable. They are not stakeholders in any
legitimate decision process.

All power to the hodlers! To paraphrase John Jay, a currency must be
governed by those who own it. Such is the creed of Bitzionism. Until
this creed is widely known and broadly agreed, Bitzion is not ready to
happen.

Technically, Bitzion must begin as a fork of the Bitcoin ledger. It
cannot work as designed unless it is born as the dominant
fork—ideally, roughly the relationship between Ethereum and Ethereum
Classic.

This requires Bitzion to achieve human reputational consensus, or
something like it, before it can even launch. At best, this would be
long, tough sledding! But afterward, the rest of the plan is actually
quite realistic.

The hodlers declare their sovereignty. To take what is theirs, they
seize the ledger. They own and will now control it. The Bitcoin ledger,
once managed by a Mos Eisley of miners, nodes, developers and
algorithms, will now be managed by a new human regime, whose exclusive
mission is to serve the hodlers of Bitcoin.

This new regime, Bitcoin’s new government, is Bitzion—still the one
and only Bitcoin. Samuel L. Jackson would like to know if you have any
questions about the new regime.

To govern Bitzion effectively as possible, the hodlers delegate their
power to a board of trustees, whom they elect. These trustees inherit
all the power of the old government.

They cannot forge transactions. They do not filter transactions. They
can and may do anything else—including creating any number of coins.
Satoshi’s constitution is dead and overthrown. The hodlers, acting
through their trustees, are now fully in the saddle.

There are two reasons for the trustees to create coins. Regular: as a
dilutive tax, to fund the normal operations of government. Special: as
an assessment to fund some one-time event—which usually means cleaning
up some altcoin.

Every revolution has a financial premise. The financial premise of this
revolution is that one of Bitzion’s first tasks will be a dramatic
reduction in mining leakage. Since Bitcoin is currently leaking about 5%
of its market cap per year via mining, it seems fair to assume that
grateful hodlers would be happy to approve a dilution tax up to a tenth
as high: 0.5% of market cap, more comparable to an expensive gold vault.

This serious budget is wholly under the control of the trustees. But the
trustees take no operational role in spending it. Instead, they delegate
all their power to a single coordinator, who coordinates all the
operations of Bitzion.

The trustees’ job is: hiring and firing the coordinator; reviewing the
coordinator’s performance; and approving regular and special tax
requests. The coordinator can request and spend regular and special
taxes. The trustees do not micromanage the coordinator. The coordinator
has no secrets from the trustees.

And the hodlers still have a job. If they become unsatisfied with the
performance of their government, they can recall it—electing new
trustees, who will choose a new coordinator. To trigger a recall, they
need a quorum defined as a percentage of the stake that voted in the
previous election.

The trustees are securely pseudonymous. No one knows who they are. No
one can pressure them. But given the legitimacy of Bitzion and its
commitment to compliance, the coordinator’s office must be completely
open, public, and transparent.

But since the coordinator works for the trustees, the coordinator is a
disposable part. All kinds of powers can pressure the coordinator. The
coordinator’s job is to either perfectly resist any such pressure, or
resign—but never to bend. A coordinator who bends will just get fired
by the trustees, a very bad and shameful outcome.

The trustees can spin up a new coordinator in another jurisdiction, or
even fall back to pseudonymous coordination. This defense works best
when the coordinator sources as much labor as possible not from
employees, but from pseudonymous contractors.

So the government of Bitzion has three teams: accountability (hodlers
and trustees); coordination (mahogany desk, press releases, lawyers,
interviews, conferences); and production (coders, writers, designers,
etc).

The accountability and production teams are pseudonymous (its members
never leak their identities, even to each other); the coordination team
is transparent (its staff are regular employees and have their pictures
on the website).

The security of trustee identity is mission-critical. The production
team can be more casual. As a labor pool, it will always experience
attrition. But any producer or trustee whose identity is revealed must
quit and cannot return, even under a new pseudonym.

The trustees’ job is: hiring and firing the coordinator; reviewing the
coordinator’s performance; and approving regular and special tax
requests. (“Tax” means dilution.) The coordinator can request and spend
regular and special taxes. The trustees do not micromanage the
coordinator. The coordinator has no secrets from the trustees.

And the hodlers still have a job. If they become unsatisfied with the
performance of their government, they can recall it—electing new
trustees, who will choose a new coordinator. To trigger a recall, they
need a quorum defined as a function of the stake that voted in the
previous election. The trustees can call a new election, too.

The sophisticated observer will recognize that this version of indirect
democracy, while owing little to historical Anglo-American political
constitutions, is kind of a blatant ripoff of Anglo-American corporate
governance.

This is unsurprising in an architecture designed to get things done. No
one says our corporations aren’t sharks. Which do you want on your
side—a shark, or a hippo? As an ex-CEO, the joint-stock company has an
antiquated, paper-belt feel in many ways. If there was anything much
more powerful, some asshole would have invented it.

Hyperdelegation

That’s the whole constitution, except for one little detail: how the
election works. Again, Bitzion is designed to have only one election
ever, unless the hodlers or the trustees get unhappy with its results
and decide to press the reset button. There are a lot of ways to elect
trustees, but here is a clever one.

A hyperelection is an election in which anyone can vote for anyone—so
long as they’re not creating a loop. At the start, you have one election
point for each coin you own. If you vote, you have zero points, and all
your points, both from your own coins and from whoever votes for you,
flow to whoever you vote for. The purpose of this design is to channel
trust, respect and loyalty upward to the most trustworthy, respectable
people.

At the end, the N leaders become the N trustees. The trustees’ votes are
weighted by the number of points they got. During the election period,
the scoreboard is public, and anyone can switch their vote at any time
(or this can be quantized into rounds). After the election, any trustee
can resign their key to anyone else.

As a total noob, the way you participate in a hyperelection is: (a)
register your wallet; (b) figure out who you know that knows the most
about Bitcoin; (c) ask that person for their pseudonym; (d) vote for
them. If you want people to be able to vote for you, also register a
pseudonym. If you have to, vote for a public candidate; but it’s always
better to just vote for a friend.

If you are a serious candidate, campaign. Publicize your pseudonym, with
or without your real name. Speak publicly to the hodlers; speak
privately with whales. During the election, if you’re not winning, vote
for someone you admire who could win.

The best precedent for this kind of social gameshow mayhem is an early
20th-century American political convention. In the end, most peoples’
votes will end up going to one of the winners, through a path of mostly
ascending respect.

Initial abnegation

But if the election leverages existing reputations by letting pseudonyms
campaign under their real names—how pseudonymous is it? It isn’t, of
course.

We fix this problem with another rule: after choosing the first
coordinator, all the initial trustees resign immediately—each giving
their key to someone they know, who they think will make an even better
trustee. So the first board of trustees after an election has two jobs.
They select both the first coordinator, and the second board.

The result of initial abnegation is a political structure in which the
second board did not choose the coordinator, so feel no attachment to
that choice; the first has no lasting power, so cannot have been
attracted to power; the second did not campaign, so were not selected by
their craving for power or status; if the second board needs to ever use
any power at all, something has already gone wrong; and the second board
is securely pseudonymous, so membership cannot confer status on the
humans behind it.

This structure is a shock absorber. Its goal is to dampen inappropriate
energy in the governance process. This bad energy could be external
pressure; it could be internal cupidity; it could even just be bad luck.
From the crooked timber of humanity nothing can be made straight, but a
good engineer can do a lot with geometry and redundancy.

Evaluation

Can anything that is not a physical state ever be impervious to state
pressure? Can all the devils of human cupidity ever be thoroughly
banished? These are easy questions. The answers are “no” and “no.”

But a stronger fortress than Bitzion will be hard to build. (The main
risk is probably that the hodlers go collectively crazy—a risk we have
to accept when we choose low, but not zero, sensitivity.)

Although adapted to the very unusual conditions of its birth and
existence, Bitzion is basically built like an early 21st-century
technology company. For all their faults and the faults of their world,
these can be very effective organizations—even at scale. Their
devotion to their shareholders is generally not in question; and their
shareholders take no actual part at all in managing them.

Let’s assume that the hodlers and trustees make good personnel choices,
and create an organization as effective and aggressive as a top
technology company. What will it do?

The Bitcoin space program

Bitzion is Bitcoin—plus a trusted central government that has no
excuse not to be as effective as any startup, if only because it is
shaped like any startup.

Let’s assume that, if the coordinator can demonstrate a use for it, the
trustees will grant the coordinator a dilution tax of up to 0.5% per
annum. At present valuations this is \$500M, which is a serious budget:
certainly enough to start a space program. Again, it is only a tenth of
the present cost of Nakamoto consensus.

The mission of the coordinator is to advance the interests of the
ledger—meaning the weighted interests of the hodlers. There is no
other restriction on the scope or nature of Bitzion’s actions. All
prudent actions designed to benefit the hodlers are authorized.

If the hodlers needed a space program, it would be proper for the
coordinator to start one. But other problems seem more pressing.

For instance, it would not be hard to spend the budget of a space
program on the full-stack Bitcoin user experience. And UX is inherently
a shallow problem. Here are three of the deep architectural challenges
Bitzion has to master.

Coordinated serialization

Nakamoto consensus is a security measure. Like any security measure, its
effectiveness must be evaluated relative to the threat model. Like any
measure, its productivity must be evaluated relative to its price. And
its price is insane.

With Bitzion as a trusted coordinator, the threat model assumes a
different shape. The Bitcoin ledger must be more trusted than the
coordinator, or even the trustees. In this design, we have not one but
two layers of security: government and decentralization.

The threat model of coordinated serialization is halfway between the
threat model of Nakamoto consensus, and the threat model of a
permissioned blockchain. It is neither completely wild, nor completely
tame.

Consider the classic threat of a 51% attack—hardly an abstract danger
when most mining is in China. 51% of all Bitcoin holders by position are
not in China or under Chinese control, so staking is more secure against
China than miner or node voting. Also, controlling miners is easier than
controlling nodes; controlling nodes is easier than controlling hodlers.

Stake attacks also exist. But a governed ledger cannot be attacked in
any such way. Not only can Bitzion reverse any such attacks, it can
punish the attacker—by banning the staked coins from the serialization
market.

Bitzion, like all governments, governs by combining the predictable
power of law with the unpredictable power of the prerogative or
exception. Judicious use of exceptional power can curate a reliable,
well-distributed volunteer serialization department.

But since the ledger must be more secure than the government, there must
be a way for the ledger to keep working if the government fails or,
worse, goes crazy. It makes sense to tie this safety mechanism to the
hodler-initiated recall mechanism.

While the serializers remain decentralized and uncoordinated volunteers,
it is hard to imagine a dying regime convincing or coercing them to
filter out a valid recall. They have no incentive to obey the regime,
and every habit and incentive to obey the recall.

Any coordinated serializers constitute an attack against the regime,
which the regime must repel assertively by one-strike banning. In
practice, it can easily verify that most of its serializers are
independent from most others. But any hint of an attempt by the regime
to coordinate its own serializers cannot go undetected, nor can it
succeed instantly; and it justifies an immediate recall.

While a new government is being elected, the ledger is in “safe mode.”
There is no exception. The volunteer serialization department operates
as usual, but without any supervision, and with higher safety
margins—which may be more expensive.

The art of solving this design problem is the art of a serialization
solution which is cheap and secure, while the government is operating
properly; and less cheap, but no less secure, if the government
disappears, fails or turns evil. In other words: without a government,
this solution must degrade into conservative proof-of-stake. This is not
an easy problem, but intuitively it feels solvable.

Bitcoin imperialism

We return to the problem of altcoins and forks. The existence of
parallel ledgers may not be a showstopper for Bitcoin—but it also
might be a showstopper.

The goal of Bitcoin is to become a standard. A standard must be certain.
Competition between coins creates uncertainty. Without these alt-dogs
nipping at its heels, Bitcoin has a stronger claim to be the one. As a
challenger, it needs all the strength it can get.

Standardizing digital money is just a flex. It demonstrates strength. It
creates strength. All healthy and growing organizations are obsessed
with their own strength. And since it must be done fairly, it it is a
chance to demonstrate the grace of kings.

So here is how Bitzion can execute a historic program of Bitcoin
imperialism, crafted in the old Silicon Valley tradition of “extend and
embrace” or “copy, acquire, kill.”

Start by evaluating all the scarcity-based tokens currently trading (not
“utility tokens,” which are virtual stonks, or “address space,” which is
virtual real estate). Classify each as worthy or worthless. Publish this
official directory and the research behind it.

A worthy altcoin is one which has made a significant contribution to the
cryptosphere—in technology, traction, or both. A worthless altcoin is
every other coin—or any new coin started after the official directory
was published.

Bitzion will never recognize these worthless ledgers. It condemns them.
They will go to zero, become deadcoins, and be forgotten. If you still
have any after the directory appears, try to sell while there are
buyers.

Bitzion recognizes the worthy ledgers and declares its intention to
merge with them—whether they like it or not. Hopefully most will like
it. The task is not easy; just doable. This makes it an excellent flex.

First, Bitcoin will incorporate all worthy innovations of these
altchains. Second, it will emulate their protocols and
semantics—including a computational emulator for computational chains.
Third, it will build a completely smooth migration path in which anyone
working with the altchain can trivially move to the real blockchain.
Most users will not even notice the difference; developers will flip a
few switches.

Now the engineering work is complete, and the marriage can be
consummated. Like a corporate merger, a blockchain merger is a
diplomatic and financial affair. It requires (a) the consent of the
altcoin holders to (b) accept some ratio of real coins per altcoin,
which can be stated as a premium or discount relative to the current
market ratio.

A corporation’s board can typically force its shareholders to accept a
merger. None of these altchains has a board; and there are no
decentralized involuntary transactions. Therefore, to merge with these
anarchic collectives, Bitzion must find, assemble, and negotiate with an
informal reputation pool of potential endorsers.

These negotiations will succeed or fail—which decides whether the
merger is friendly or hostile. Even a friendly merger cannot be
enforced, though its chances are better. If negotiations cannot set a
fair price, Bitzion will set one. Generally the hostile price is lower
than the friendly price—pour encourager les autres.

Then, altcoin holders have a short period to make a decision. They can
either sell their altcoins for real coins at the final ratio, or grimly
hold on. If they sell, they get new but genuine coins—which the
coordinator has minted with the trustees’ approval..

In exchange, the coordinator gets all their altcoins. After the merger
and migration, the coordinator sells these altcoins—using the
revenues, if any, to purchase real coins.

If even a low double-digit percentage of altcoins gets tendered, selling
them blows through the altchain’s whole buy-order book like a firehose
up a salmon’s asshole, and takes the market price of this legacy chain
to zero, where it should stay: a deadcoin.

So grimly holding on is kind of a bad move. Are these the tactics of
Genghis Khan? Yeah, kind of. Do you know more about imperialism than
Genghis Khan?

Bitzion starts this imperialist program on the bottom end of the worthy
list—taking down its feeblest competitors first. As these involuntary
mergers succeed, reducing any foolish holdouts to pyramids of 256-bit
skulls, the empire’s negotiating position grows stronger and stronger.
Oderint dum metuant!

The Bitcoin intelligence agency

The most important problem of any regime is security. For a blockchain,
security is serialization plus imperialism. The second most important
problem is intelligence.

This isn’t James Bond stuff, of course. Intelligence in the modern sense
of the term just means collection and analysis of accurate and official
information. It doesn’t require any kind of secret volcano island
base—although that would certainly be cool. There are no martinis and
no silencers. Yes, we do like cats.

Why do blockchains need official government intelligence? Intelligence
is data plus analysis. Analysis is great—but the first thing
blockchains need is data. They even have a word for an official data
stream: an oracle.

The Sun has been around for four billion years. On the full power of a
computational blockchain plus an official data stream, the Sun has never
yet looked. In theory, it is possible to build prediction markets with
oracle reputation as a second moving part. In practice these weird,
wobbly markets suck.

Without this concrete and certain connection to official facts, a
computational chain simply cannot think about the real world. A
blockchain which can compute with facts about the real world is a
different class of tool. “Prediction markets” does not cover the
potential power of this architecture.

Of course, the difference between data and analysis is not always clear.
Once Bitzion becomes comfortable with its right to curate official data,
it will inevitably find itself considering the next step of composing
official analysis.

Data feeds algorithms. Analysis feeds humans. The algorithms on a
blockchain can only use data. But those algorithms are ultimately
working for humans; and humans can use both data and analysis.

The mission of Bitzion is to serve the hodlers. Authoritative
open-source real-world intelligence, also known as “news,” is often
suspect—less the data, more the analysis. If the only narrative
available to the hodlers is this “news,” and the “news” contains
systematic misjudgments, the hodlers must be likely to make systematic
mistakes.

So Bitzion’s intelligence agency starts by collecting real-world
facts—which helps blockchain programs think about the real world.
Finding its feet with this easy work, it continues by constructing
real-world narratives—which helps blockchain users think about the
real world.

Again, these are exactly the nominal functions of any 21st-century
intelligence agency. All these agencies still have some kind of weird
1950s James Bond historical mystique. The James Bond era was real; if it
even lasted into the ’50s, that would be surprising.

Today the intel agencies are mostly just government think tanks, whose
wisest wisdom is not much different from the wisdom of the press and the
universities. The best test of whether they matter is what would happen
if they disappeared tomorrow, which is absolutely nothing.

Yet their nominal function—research and analysis—remains real and
relevant. Nothing, moreover, could be more legitimate. This proposed
Bitcoin intelligence agency is, in the words of our peach-haired
President, “very legal and very cool.”

Of course, Bitzion keeps no secrets. All this intelligence product is
public. And like the rest of Bitzion, it is independent of real-world
power.

Needless to say, raw input from “the news” or even “science,” even from
the most “reliable” or “respected” sources, is never incorporated
directly and without a clear positive argument into the output of the
analysis directorate.

Rather, its mission is to understand the world anew from scratch.
Everything “known” is an invaluable resource, of course. But just a
resource. How do we “know” this? Imagine trying to recreate Wikipedia
with no “reliable” sources—but instead, the conviction that all
information from the 20th century is unreliable until shown otherwise.

So is all external information. While an intelligence agency wants to
hear everything, its mission is to think entirely for itself. Nothing in
its raw input stream is true until the agency has some clear positive
reason to consider it true. “Always trust content from Microsoft
Corporation” is not a valid form of reason.

This question need not be asked story by story or paper by paper. It is
best asked field by field, department by department, agency by agency.
If the old 20th-century truth is a blueberry pie, we can pick out the
moldy parts slice by slice, not berry by berry. The whole pie is never
moldy—however old, there are always good parts in a pie. Pure math,
for instance, remains sweet and delicious with only a very mild
fermented bite. Then again, pure math did pretty well in the USSR.

At this point we imagine a genuine conflict between virtual and physical
power—but this is probably a good stopping point for today. Once
Bitcoin has gone so far toward making a nation that it has made an
intelligence agency, the point seems proven.

Both necessary and impossible

Of course, the CJNG must feel the same way about its little panzer
parade— and that actually happened.

Anyone who has actually read all this perhaps feels the same intuitive
sense as me: that Bitzion theoretically should happen, but actually
can’t happen. It is both necessary, and impossible; both fantastic, and
pragmatic.

I have put a lot of words into why Bitzion is necessary. I already
warned you that it is impossible. If it becomes possible, it will not
become possible easily.

Our theoretical interest, as political engineers, in what should be
done, remains legitimate. But a thought-experiment and a practical
conspiracy are two things. The conversation is not over without talking
about why it cannot happen. So let’s do that.

It’s simple. Bitzion cannot happen because vanity is poison and everyone
is poisoned.

Recently, Marc Andreessen wrote a much-cited essay which asked everyone
to build. Everyone should build. Not everyone should be an architect.

The trouble with asking the whole world to build is that this world is
so poisoned by vanity that it can only hear one message: be an
architect. Almost all these people should be learning to put one brick
on top of another. But once a thousand architects have argued about how
to build the house, there is often not a bricklayer to be found.

Having done both—metaphorically speaking—I will tell you
straight-up: it is more honorable to be a bricklayer. Architecture is a
gift. Masonry is a craft. Never trust anyone who has a gift but no
craft. If you have a gift, pursue your craft. Only fate can make you an
architect; but fate seems to like a man who can lay a brick.

To lay a brick is to pursue the path laid down for you; it is to follow.
The mason is not an artist but an artisan. For an artist, the definition
of perfect is broad; for an artisan, it is narrow. The same amount of
effort and achievement can go into either.

Yet the architect, who is an artist, leads; the mason, who is an
artisan, follows. The wall needs to go where it is supposed to go. It
needs to look like what it is supposed to look like. Yet in the hands of
an ordinary mason, it will be okay; in the hands of a fine mason, it
will be striking; anyone can tell the difference and no one can say
quite why.

While the society that forces an artist to be an artisan has failed, so
has the society that forces an artisan to be an artist—or that exalt
the artist above the artisan. The latter is simply a society without
artisans—which is a society that cannot build. Oops.

(In fact it is a curse to be a real artist or scientist. These people
should be derided and shunned. It’s an admission against interest—but
that way, they’ll do their best work.)

This is where Bitzion fails. Bitzion is obvious. Bitzion would already
have happened— in a world where people thought about collective action
not in terms of leading, but in terms of joining and following—where
they knew how to be masons, not architects.

Not that anyone knows how to be an architect, either. Being an architect
means making plans which masons follow. There are no masons; so there
are no architects. No one has any experience in the actual profession of
architecture. There are just dreamers who try to build beautiful dreams,
by themselves, sloppily. Many houses are started; few are finished; many
of these fall down. But the rest are beautiful.

This metaphor is just a metaphor. It could describe many careers and
professions. In all such cases, there is one factor in common: vanity is
poison.

In the case of Bitzion, vanity manifests itself as an inability to
trust. To the average hodler, the thought of delegating collective
control to the board of trustees—even through this clever mechanism,
by which no one’s vote is lost—is horrifying.

Even the thought of trusting the hodlerate collectively would horrify
most hodlers. Literal collectivism? Why, isn’t Bitcoin supposed to be
the exact opposite of that?

But worst of all—hyperdelegation, or any sort of delegation, or
election, or indirect democracy in general, means you are giving away
your power. Why vote, when you could rule? It’s going from weed to
meth—having ruled, it will never be enough to just vote.

For example, when you vote, you vote every four years—for four years,
except one day, you have zero power. When you rule, you get to rule
every day of the week. And even worse: when you vote, you follow. You
literally submit to another person. You almost feel your genitals
shrinking.

To be a mason is to say to some other human or humans: I will obey you,
I will follow your directions, I will put the bricks how and where you
tell me to put them, I will not create my own whimsical Art Nouveau
motifs.

Of course, to join together and follow a common leader, or leadership,
or government, or constitution, is the fundamental act which creates all
regimes. This act is necessary. Without such an act of common consent,
no political structure can be created.

But one and only one word of this recipe is utterly impossible. That
word is follow. No one wants to follow. Worse: no one even knows how to
follow.

The practical consequence of this deficit is an inability to trust. The
project of Bitzion is impossible because it depends on taking a hodler
community dedicated, X-Files style, to trusting no one; and moving it to
trusting each other; and implementing this by trusting a roomful of
people, who then trust a single person.

To a true Bitcoiner, this seems completely insane. It is also how every
company in the world, major or minor, works—and Bitcoin has the market
cap of a major company. And in any major company, the amount of time the
shareholders spend even caring about the governance of the company, let
alone trying to change it, is epsilon.

If they are not delegating unconditional trust, they are delegating
something close. They have no plan at all for running the company
themselves, or even interceding in its decision loop. There are always
exceptions; but this is the normal operating loop. The shareholders seem
to trust their direct representatives, the directors, no less than they
trust the CEO selected by those directors. This is an amazing amount of
trust—which can only exist because the joint-stock company was
invented long ago.

Everything big that works is worked by these companies. They are clearly
the most effective way to operate at scale. Yet we are incapable of
recreating this level of trust in a new context. The joint-stock company
design survives because it is grandfathered—and because it works so
well. Nonetheless, a “capitalist” government is no more than one which
allows joint-stock companies—so there remains a huge target on its
back.

And yet, the idea of the hodlers of Bitcoin collectively agreeing to
turn themselves into the sharehodlers of Bitzion, which is the same as
Bitcoin except that it governs itself rather than being governed by
math, seems at a certain level preposterous—because it goes from
trusting math, to trusting humans. And trusting humans is following. And
following is servitude—which was exactly what we were trying to escape
from, right?

In practice Bitcoiners have to trust humans anyway; only a government
which is both legitimate and human can be proactive, which means only
such a government can be effective; and government by hash is incredibly
expensive. And yet Bitcoin remains the leader—proving the power of
monetary incumbency.

Bitcoin can’t do a 180 and treat decentralization not as an end, but as
a means—to the real goal, independence. This would open the door to
independence through a wholly different means, pseudonymous
centralization. It seems like a minor, obvious change. But it would
require Bitcoin to entirely change its mind—or at least, its religion.

Bitcoin is a cult, for better or worse. Cults don’t change their
collective minds from the bottom up. Bitcoin is a cult with an
accidental and mathematical government. Bitzion is the exact same cult,
with an conscious and human government.

An altcoin or a fork with this design could still be organized; but it
would still be an altcoin. Please don’t try this. Bitzion has to be all
of Bitcoin, or most of it; and if didn’t ask you to follow, it might
have a chance at that. But it exists to ask you to follow.

Honor still demands honesty. In a society where everyone is too good to
lay bricks, no one can build a house. It’s a hard problem, and it won’t
get solved here. But solving the problem has to start by talking about
it.

Consider the irony. This ancient human talent—the reflex for
unconditional loyalty—which could make Bitcoin a thousand times more
powerful—is, for a bunch of peasant narco-soldiers from Jalisco, as
easy as falling off a log.

Yet you might as well ask the hodlers, our fellow autistic rationalist
superbrains, who whether they like it or not are citizens of this new
thing with a B, to dunk a basketball. Our 20th-century prejudice is that
intelligence is power. Sometimes it is weakness; and all we have to
fight this weakness is intelligence itself.