Business
jo_12
Cryptocracy-TheQuesttoReplacePoliticswithTechnology.pdf
When the dao prevails in the world, the common people do not discuss gover nance.
— The Analects of Confucius 16.21
“The whole country was in the hands of a few persons, and if the tenants
failed to pay their rent they were liable to be haled into slavery, and their
children with them.”2 Aristotle in his Athenian Constitution describes how
the ancient city state was once ruled by ruthless oligarchs. Government
administrators were perceived as corrupt and untrustworthy, offering no
recourse to the oppressed. The situation became untenable:
Since . . . the many were in slavery to the few, the people rose against the upper class. The strife was keen, and for a long time the two parties were ranged in hostile camps against one another. . . .
A surprising resolution to the conflict was found in the appointment of a
poet to devise a new system of government for the country:
at last, by common consent, they appointed Solon to be mediator and Archon, and committed the whole constitution to his hands.3
Solon was a poet but also a competent statesman. He sang a poem and
then set about designing a better system. Instead of trying to make govern-
ment administrators more trustworthy, he took a different approach: he
8 CRYPTOCRACY: THE QUEST TO REPLACE POLITICS WITH TECHNOLOGY
This is a portion of the eBook at doi:10.7551/mitpress/14219.001.0001
Downloaded from http://direct.mit.edu/books/book/chapter-pdf/2423278/c005400_9780262371094.pdf by CSU: San Francisco user on 26 August 2024
132 CHAPteR 8
wanted to make trustworthiness matter less. Achieving this involved a
machine, called the kleroterion, or “allotment machine.”4
The kleroterion was about the height of a man and built around a rect-
angular slab of stone. Carved into the face of the slab was a matrix of slots,
about ten columns across and fifty rows down. Inserted into the slots were
bronze plates bearing the names of the people who had shown up to the
machine that day. Things clicked into action when white and black balls
were dropped into a funnel attached to the side of the slab. The balls tum-
bled and mixed inside the machine, until a mechanism released them, row
by row. A black ball meant that the people whose names were on that row
were sent home. A white ball meant that they were appointed as govern-
ment administrators.
Using the kleroterion, random people were selected to serve as govern-
ment administrators in ancient Athens.5 Magistrates were appointed in
this fashion annually. Judges were reselected every morning. Each legal
case in Athens was heard by several of these randomly selected judges, who
acted as checks on each other. As long as the majority were honest, a few
corrupt officials could not abuse their power. Individual trustworthiness
would not matter. To incentivize participation, each appointee received a
reward from the public purse.
Having designed his revolutionary system of government and seen it
off to a start, Solon set off on a journey to Egypt. He disappeared from
public life and let others carry on the project.
PROBLEM OF TRUST
Satoshi Nakamoto was a skilled if somewhat old- fashioned programmer.6
On his online profile he claimed to be born in 1975 and live in Japan, but
this was probably just an online persona that he, she, or they had cre-
ated for themselves. “Nakamoto” wrote messages in precise British English,
cited the London- based Times newspaper, and was active mostly during
British daytime hours.
Whoever he was, Nakamoto’s messages suggested that he was disap-
pointed with how the digital revolution had turned out. Cybervisionaries
like John Barlow had imagined that the Internet would give rise to social
order beyond the reach of governments and powerful corporations.7 Yet
This is a portion of the eBook at doi:10.7551/mitpress/14219.001.0001
Downloaded from http://direct.mit.edu/books/book/chapter-pdf/2423278/c005400_9780262371094.pdf by CSU: San Francisco user on 26 August 2024
CRyPtoCRACy 133
by the late 2000s, it was clear that government was not going anywhere
and that the Internet was giving rise to corporations that, if anything,
were even more powerful than before. In particular, Nakamoto was both-
ered by how people still had to rely on powerful and opaque financial
institutions to manage their finances:
The root problem with conventional currency is all the trust that’s required to make it work. The central bank must be trusted not to debase the currency. . . . Banks must be trusted to hold our money and transfer it electronically. . . . 8
Commerce on the Internet has come to rely almost exclusively on financial institutions serving as trusted third parties to process electronic payments.9
Many people at the time had similar sentiments. In 2008, the world was
reeling from the effects of the great financial crisis. Due to inept gov-
ernment and selfish corporations that had misled and defrauded their
customers, many people had lost their jobs, their savings, and even their
homes. Thousands protested on Wall Street, demanding a greater say in
how these institutions were being run.
But Nakamoto was not interested in making the institutions more
democratic. Instead, he wanted to resuscitate the Barlowian dream of a dig-
ital social order that wouldn’t need such institutions in the first place— no
bureaucrats, no politicians who inevitably betrayed their electorates’ trust,
no elections rigged by corporations, no corporate overlords. Nakamoto still
thought that such a social order could be created with technology— and in
particular, with cryptographic technology.
Nakamoto was not the first to believe in the power of cryptography
to achieve such goals. A whole subculture of programmers calling them-
selves “cypherpunks” and “crypto- anarchists” had been pursuing political
liberation through cryptography for almost two decades.10 Derived from
the ancient Greek words kryptós (hidden, secret) and gráphein (to write),
cryptography is the millennia- old craft of creating and deciphering secret
messages. The advent of personal computing gave the discipline a huge
boost, and the cypherpunks’ mailing list at one point reached thousands
of subscribers.11
Cypherpunks’ goal was to create infrastructures that could not be con-
trolled by authorities, whether state or corporate. So far, they had success-
fully built anonymous communication platforms that allowed messages
This is a portion of the eBook at doi:10.7551/mitpress/14219.001.0001
Downloaded from http://direct.mit.edu/books/book/chapter-pdf/2423278/c005400_9780262371094.pdf by CSU: San Francisco user on 26 August 2024
134 CHAPteR 8
to be exchanged beyond authorities’ oversight. But after years of work,
they still had not succeeded in building viable payment platforms. As a
result, their enclaves remained all talk and no business. “We must come
together and create systems which allow anonymous transactions to take
place,” “A Cypherpunk’s Manifesto” had urged already back in 1988.12
Twenty years later, the goal seemed as elusive as ever, and the movement
was losing steam.
To understand what the movement was trying to achieve, consider
for instance the functions of a conventional bank. A bank ensures that
whenever someone wants to make a payment, that person actually has
enough credit on their account. If the account balance is insufficient, the
bank stops the transaction from happening. The bank’s oversight ensures
that the same money cannot be spent twice and that account holders
cannot create money out of thin air. In this and other ways, the finan-
cial system creates order and makes economic exchange possible between
people who entrust their funds to it. But financial institutions can also
abuse that trust— refuse valid transactions, hold monies hostage, or rig
rules to favor insiders, for instance. Trust means belief in someone’s good
intentions despite an absence of guarantees, so risk of abuse is inherent to
it. It boils down to the age- old problem of political science that troubled
ancient Athenians, too: the authorities protect us, but who will protect us
from the authorities? How can we hold power to account? Cypherpunks
and crypto- anarchists called it the “Problem of Trust,” and they wanted
to solve it with technology.13
In 1990s, entrepreneurs launched new digital payment platforms that
challenged banks’ monopoly on mediating payments. Peter Thiel, Elon
Musk, and their cofounders started what eventually became PayPal, the
most successful of these ventures. But PayPal’s administrators imposed fees
and policies that many users felt were arbitrary and opaque. There was
nowhere to appeal if the platform froze a merchant’s account and put them
out of business. PayPal broke new ground in facilitating transactions over
the Internet, but it solved nothing when it came to the problem of trust.
Some digital payment platforms tried to adopt more liberal approaches.
A platform called E- gold asked few questions of its users and rarely policed
transactions. It quickly attracted criminal money, and the US government
shut it down and arrested its owners.
This is a portion of the eBook at doi:10.7551/mitpress/14219.001.0001
Downloaded from http://direct.mit.edu/books/book/chapter-pdf/2423278/c005400_9780262371094.pdf by CSU: San Francisco user on 26 August 2024
CRyPtoCRACy 135
Nakamoto surmised that the problem with these platforms was that
they still placed power in the hands of a central authority whom users
had to trust:
A lot of people automatically dismiss e- currency as a lost cause because of all the companies that failed since the 1990s. I hope it’s obvious it was only the centrally controlled nature of those systems that doomed them.14
A trusted central party could abuse its power, as platform companies often
did. At the same time it was also vulnerable to government take- down. To
avoid these pitfalls, Nakamoto wanted to create a “trustless” platform— one
in which the trusted authority was replaced with technological certainty:
What is needed is an electronic payment system based on cryptographic proof instead of trust, allowing any two willing parties to transact directly with each other without the need for a trusted third party.15
Instead of having a central entity like PayPal mediate people’s pay-
ments, Nakamoto wanted people to be able to send payments directly
to each other. To make this happen, every participant in Nakamoto’s
scheme would run special peer- to- peer “banking software” on their com-
puters, which communicated directly with other participants’ computers.
Nakamoto took inspiration from peer- to- peer (P2P) file sharing:
Governments are good at cutting off the heads of a centrally controlled net- works like Napster, but pure P2P networks like Gnutella and Tor seem to be holding their own.16
The ledgers of this peer- to- peer payment platform would not be held
in any central database but as parallel copies on every user’s computer.
Nakamoto called it a “decentralized” platform.
THWARTING THE SYBIL ATTACK
How would such a decentralized platform ensure that people spent only their
own monies? PayPal authenticated users by asking them to log in with
their usernames and passwords. But in a peer- to- peer system, broadcasting
your username and password to every other user was hardly a good idea.
Ancient Athenians sometimes used pieces of ceramic to authenticate
themselves. A flat piece of ceramic with a name or sign was broken in
half, leaving an irregular edge on both halves. One half of the ceramic
This is a portion of the eBook at doi:10.7551/mitpress/14219.001.0001
Downloaded from http://direct.mit.edu/books/book/chapter-pdf/2423278/c005400_9780262371094.pdf by CSU: San Francisco user on 26 August 2024
136 CHAPteR 8
was given to an administrator, and the other half taken by a person who
later needed to authenticate themselves. That person could then later
prove who they were by demonstrating that they and only they held a
piece of ceramic that perfectly matched with the fragment held by the
administration. Unlike a password, the fragment could not be copied, not
even by someone who possessed the other half.
Crypto- anarchists before Nakamoto had figured out that a similar tech-
nique, called digital signing, could be used to authenticate users in a peer- to-
peer system. Instead of ceramics, the technique relied on specially devised
pairs of numbers, known as keys. One of the keys was taken and held pri-
vately by the account holder, and the other key was used as the account
number and broadcast to everyone else. Like two halves of a ceramic, the
two keys formed a perfect mathematical fit; they could be used by the
account holder to prove their ownership of the account to other people,
without giving away anything that would allow others to imitate them.17
However, digital signing had not allowed crypto- anarchists to create
truly trustless payment platforms, because a trusted authority was still
needed for another reason— to keep track of account balances and check
that monies had not already been spent. Thus, by late 2000s, digital sig-
natures were widely used, but they were used by banks, payment compa-
nies, and other trusted digital platforms.
Nakamoto had a new idea: the responsibility for checking balances
could circulate randomly between users, a little like how administrator
posts circulated randomly between citizens in ancient Athens. Where
Athenians used the kleroterion to rotate administrators every twenty- four
hours, Nakamoto’s scheme used an algorithm to rotate the administrator
approximately every ten minutes.
The job of the administrator in Nakamoto’s system was to go through
recently issued payment instructions, check that they were valid, and
collate them into a record known as a block— an official record of transac-
tions that could be used to determine who owned what in the system. Of
course, the administrator would not have to check transactions by hand:
all the work would be done automatically by the peer- to- peer “banking
software” running on their computer.
After approximately ten minutes, the next randomly appointed admin-
istrator would take over, double check the previous block of records, and
This is a portion of the eBook at doi:10.7551/mitpress/14219.001.0001
Downloaded from http://direct.mit.edu/books/book/chapter-pdf/2423278/c005400_9780262371094.pdf by CSU: San Francisco user on 26 August 2024
CRyPtoCRACy 137
append their own block to it, forming a chain of blocks. Just like in
ancient Athens, this constant circulation of responsibility meant that the
administration would be extremely difficult to corrupt. Together the users
would be as powerful as a bank, but individually none would wield power
sufficient to coerce another. As long as a majority of the peers remained
honest, the platform could maintain orderly records without any single
trusted authority. Belief in good intentions was replaced with technologi-
cal certainty. The problem of trust appeared to be solved:
With e- currency based on cryptographic proof, without the need to trust a third party middleman, money can be secure and transactions effortless.18
However, a significant problem remained. What if an attacker created
puppet accounts until their numbers overwhelmed the legitimate users?
It was not difficult to create lots of new digital personas for oneself, espe-
cially among crypto- anarchists who swore by privacy and anonymity. The
randomly chosen administrator would then in reality end up being the
same person again and again, undermining the system’s supposed lack of
reliance on any single party. This so- called Sybil attack— named after the
Greek pseudonym of a woman who supposedly possessed sixteen differ-
ent personalities— had stumped earlier crypto- anarchists.
Poet Solon had faced a somewhat analogous design problem. In Ath-
ens, a randomly selected administration could in theory have been taken
over by people from a rival city state. Rivals could have suddenly shown
up in the morning in great numbers and stuffed the kleroterion with
their own nameplates. Once appointed to a majority of the city’s admin-
istrative posts, they could have wreaked havoc.
Solon prevented such attacks by limiting eligibility to men who could
prove that they owned property in Athens. Candidates’ plates were sorted
into the machine’s columns in accordance with their wealth, so that each
row represented a spectrum of men from the richest to less well off. For
those with no property, the machine simply had no column (nor did it
have columns for women or for slaves).
Nakamoto’s defense against a Sybil attack was somewhat similar: his
scheme required would- be administrators to prove that they owned a CPU.
A CPU or central processing unit is the part of a computer that makes
calculations. Anyone wishing to have a shot at being selected as the next
This is a portion of the eBook at doi:10.7551/mitpress/14219.001.0001
Downloaded from http://direct.mit.edu/books/book/chapter-pdf/2423278/c005400_9780262371094.pdf by CSU: San Francisco user on 26 August 2024
138 CHAPteR 8
administrator in Nakamoto’s scheme had to make their computer’s CPU
try to guess a number that would solve an otherwise meaningless crypto-
graphic puzzle. The first participant whose CPU guessed the correct num-
ber became the administrator for the next ten- minute block. Although
anyone could create as many online personas as they liked, if the personas
shared the same computer, their combined likelihood of being appointed
would still be no greater than the individual’s alone, so they would gain
nothing from it. Sybil was thwarted.
Nakamoto didn’t invent the technique of requiring users to spend
CPU cycles on guessing numbers. It was known among crypto- anarchists
as proof- of- work, work being the cycles expended and proof being the cor-
rect guess. But Nakamoto’s idea of using the technique to select a ran-
domly rotating recordkeeper appeared to be a breakthrough. After years
of frustration, the crypto- anarchists’ dream of a reliable payment plat-
form without a trusted authority suddenly seemed within reach. “Every-
thing is based on crypto proof instead of trust,” Nakamoto summarized
his invention.19
THE MOST DANGEROUS PROJECT
On October 31, 2008, Nakamoto announced his invention to the world:
From: <satoshi@vistomail . com> To: The Cryptography Mailing List <cryptography@metzdowd . com> Subject: Bitcoin P2P e- cash paper
I’ve been working on a new electronic cash system that’s fully peer- to- peer, with no trusted third party.
The paper is available at: http:// www . bitcoin . org / bitcoin . pdf . . .
— Satoshi Nakamoto
Two months later, Nakamoto released version 0.1 of his software. It was
the peer- to- peer “banking program” that people would run on their com-
puters to join the network, issue transactions, and— if they so wished—
spend CPU cycles to compete for a spot as the administrator. He set up
the network’s first node and recorded the first transactions. Into the first
transaction record block, he encoded a message— a newspaper reference
that dated the record but also mocked the institutions that his system was
set to challenge:
This is a portion of the eBook at doi:10.7551/mitpress/14219.001.0001
Downloaded from http://direct.mit.edu/books/book/chapter-pdf/2423278/c005400_9780262371094.pdf by CSU: San Francisco user on 26 August 2024
CRyPtoCRACy 139
The Times 03/Jan/2009 Chancellor on brink of second bailout for banks
Nakamoto received a lukewarm response at first. After decades of fail-
ure, many crypto- anarchists had grown pessimistic about the prospects of
a truly trustless digital payment system. But a handful joined their comput-
ers to his network and began to play around with the platform. Unlike Pay-
Pal, the platform couldn’t be used to issue payments in US dollars or any
other national currency. Instead, the numbers recorded into the chain of
blocks represented a new currency unit— bitcoin. The virtual coins weren’t
worth anything as such. They were just tokens. Users sent them back and
forth just to test the system. Nakamoto released updates to the software,
fixing bugs and adding features. He also set up a mailing list and an online
discussion forum for people interested in the project.
I keep a list of all unresolved bugs I’ve seen on the forum. . . . This isn’t the kind of software where we can leave so many unresolved bugs that we need a tracker for them.20
After enthusiasts had been testing and tinkering with the platform for
over a year, Bitcoin still hadn’t seen any real use as a payment system. Like
any platform, it faced a chicken- and- egg problem: How to attract consum-
ers when no businesses accepted bitcoin? How to attract businesses when
no consumers used it?21
Thanks to the financial crisis, trust in established institutions was at
a low point. Many people were eager for change. Some small businesses
began to experiment with the new digital currency, which was said to offer
independence from the old regime. A vegan café near where I lived adver-
tised a soy- based cheeseburger meal for one bitcoin. An online merchant
began to sell alpaca socks in bitcoin. The Electronic Frontier Foundation
began to accept donations in the currency.
But cafés, merchants, and foundations still had to pay their suppliers
and employees in local currency. Perhaps in the future, they might be
able to pay suppliers in bitcoin, if it grew into a widely accepted currency.
That was what Nakamoto was hoping for. But for now, it was still neces-
sary to convert the virtual coins into a national currency. This was for-
tunately possible on exchanges— emerging online trading sites on which
people began to buy and sell bitcoins for dollars, euros, pounds, and yen.
This is a portion of the eBook at doi:10.7551/mitpress/14219.001.0001
Downloaded from http://direct.mit.edu/books/book/chapter-pdf/2423278/c005400_9780262371094.pdf by CSU: San Francisco user on 26 August 2024
140 CHAPteR 8
As word of the new anti- authority payment platform spread, more peo-
ple became interested. Bloggers took notice. American tech investor Jason
Calacanis wrote that Bitcoin was “the most dangerous open- source proj-
ect ever created . . . unstoppable without end- user prosecution.”22 Swedish
libertarian activist Rick Falkvinge explained that “you can transfer any
amount anywhere instantly without any authority knowing or interfer-
ing” and announced he was “putting all my savings into Bitcoin.”23 Then
media outlets picked up the story. Wired magazine called Bitcoin “math-
based money” that was immune to human politics.24 A New York Times
Magazine cover depicted a dollar bill dissolving into pixels, overlaid by the
words “In Code We Trust.”25 Bitcoin’s exchange rate soared.
With his system successfully inaugurated, Satoshi Nakamoto began to
step back from the project. His forum posts became less and less frequent
until they stopped completely. Like the poet Solon, the pseudonymous
programmer disappeared from the public stage and entered into legend.
A BUG IN THE MACHINE
Bitcoin’s success inspired others to initiate similar projects. Some simply
copied Nakamoto’s source code, changed a few variables, and launched a
competing platform with a new name and currency unit. These platforms
and their tokens came to be known collectively as cryptocurrencies.
Others sought to go further. Brilliant Russian Canadian programmer
Vitalik Buterin and his collaborators created a system called Ethereum. Like
Bitcoin, it used a proof- of- work algorithm to randomly appoint compu-
tational recordkeepers who strung blocks of transactions together into
an official record. The name of its currency unit was ether. But instead of
mere payment transactions, its record could also contain smart contracts—
programs specifying that a payment should be carried out only when cer-
tain conditions were met, for instance. Where legal contracts are written in
English and executed by lawyers and courts, smart contracts were written
in computer code and executed by the Ethereum peer- to- peer network.
Of course, automated programs were already performing conditional
transactions everywhere from financial markets and company payrolls to
Google Ads and the Amazon Marketplace. The difference was that the
Ethereum platform promised to run such code “trustlessly”: if parties
This is a portion of the eBook at doi:10.7551/mitpress/14219.001.0001
Downloaded from http://direct.mit.edu/books/book/chapter-pdf/2423278/c005400_9780262371094.pdf by CSU: San Francisco user on 26 August 2024
CRyPtoCRACy 141
submitted a contract to it, it would execute that contract exactly as written
and nothing else. It would be guaranteed not to overstep its authority and
abuse its power. There would not be “any possibility of downtime, censor-
ship, fraud, or third- party interference,” according to the platform’s website;
contracts would be “unstoppable.”26 Users would not have to take Buterin’s
word for it: certainty would be built into the technology itself, using the
techniques that Nakamoto had pioneered. It would solve the problem of
trust not just for payments but for all kinds of economic interactions.
This promise carried awesome political ramifications. Throughout his-
tory, the state and other formal institutions had played an indispensable
role in economic growth by enforcing contracts and property rights. If the
Internet had recently diminished the state’s role, it was only by replac-
ing it with private state- like authorities.27 Our reliance on authorities for
order continued to leave us vulnerable should they turn against us. Mil-
lennia of political science had not delivered any definite answer to this
fundamental problem. Now Ethereum promised to solve it and deliver
formal institutions’ benefits without their risks. “Can’t be evil > don’t be
evil,” summarized blockchain visionary Chris Dixon.28 Peter Thiel paid
Buterin $100,000 to drop out of college and focus on Ethereum full-time.
One year from the network’s launch, the reality was somewhat less
awe- inspiring: most of the popular smart contracts on Ethereum were
gambling machines, Ponzi schemes, and other unimpressive undertak-
ings. But one shining example of the platform’s potential was The Distrib-
uted Autonomous Organization (DAO), a complex set of smart contracts
initiated in April 2016. According to its website, it was a “new breed of
human organization never before attempted . . . borne from immutable,
unstoppable, and irrefutable computer code.”29
Participants could deposit money into The DAO in exchange for vot-
ing rights that determined how the code would invest its funds. Any prof-
its would be credited to the participants’ virtual accounts. The DAO thus
resembled an investor- directed venture capital fund, except that it was
not incorporated under the laws of any state: its by- laws were expressed
in computer code and administered only by the peer- to- peer network
whose nodes now dotted the world. It was, according to its German cre-
ators, “existing simultaneously nowhere and everywhere and operating
solely with the steadfast iron will of unstoppable code.”30 There were
This is a portion of the eBook at doi:10.7551/mitpress/14219.001.0001
Downloaded from http://direct.mit.edu/books/book/chapter-pdf/2423278/c005400_9780262371094.pdf by CSU: San Francisco user on 26 August 2024
142 CHAPteR 8
human- readable explanations of how The DAO worked, but its creators
stressed that the real rules were those expressed in its code:
Any and all explanatory terms or descriptions are merely offered for educational purposes and do not supersede or modify the express terms of The DAO’s code set forth on the blockchain.31
Crypto- anarchists and tech journalists were enthralled. A TechCrunch
story described The DAO as “a paradigm shift in the very idea of economic
organization.”32 In its first month, the fund attracted over $150 million
worth of investment from over eleven thousand people.33
It is a well- established tenet in software engineering that software is
never perfect: despite programmers’ best efforts, defects or “bugs” are
guaranteed to remain in almost any code. A metric used to measure soft-
ware quality is defects per thousand lines of code (or KLOC). According to
one industry estimate, freshly written code typically contains around ten
to fifty defects per KLOC, while fully tested code usually contains up to
0.5 defects per KLOC.34 A study of popular open source software packages
reported an average of 0.69 defects per KLOC.35 Sometimes critical bugs
are not discovered until years or even decades later.
At just over two thousand lines of code, The DAO was small by software
project standards. It was also carefully vetted before release. But statisti-
cally speaking, it was still likely to contain bugs. And indeed, several were
discovered within weeks. On June 17, someone began to exploit them. “I
think TheDAO is getting drained right now,” wrote a pseudonymous user
on Ethereum’s discussion forum.36 Vitalik Buterin and many others showed
up. But they could only watch as cryptocurrency gradually disappeared
from the fund, moved out in tranches. In the end, the hacker managed to
siphon out around a third of the fund’s treasury— about $50 million worth
of ether.
SOFTWARE UPDATE
As news of the hack spread, The DAO’s investors were shocked. Many
took to the forum to demand recourse. But from the Ethereum platform’s
point of view, no rules had actually been broken. The alleged hacker had
simply made use of features present in The DAO’s code to withdraw funds
for themselves. Whether The DAO’s creators had put those features there
on purpose or not was not something that the automated platform was
This is a portion of the eBook at doi:10.7551/mitpress/14219.001.0001
Downloaded from http://direct.mit.edu/books/book/chapter-pdf/2423278/c005400_9780262371094.pdf by CSU: San Francisco user on 26 August 2024
CRyPtoCRACy 143
in a position to judge. The platform’s job was simply to execute the code
as written; any bugs were in the eye of the beholder. An anonymous mes-
sage purporting to be from the hacker emphasized this point:
I have carefully examined the code of The DAO and decided to participate after finding the feature where splitting is rewarded with additional ether. I have made use of this feature and have rightfully claimed 3,641,694 ether, and would like to thank the DAO for this reward. . . . I am disappointed by those who are characterizing the use of this intentional feature as “theft.” I am making use of this explicitly coded feature as per the smart contract terms. . . . Yours truly, “The Attacker”37
The situation was a catastrophe for The DAO and its investors but also
for the entire Ethereum platform. The Distributed Autonomous Organiza-
tion was the platform’s model application and leading media case study.
Around 15 percent of all ether in circulation was by this time invested in
The DAO. If investors simply lost their funds, faith in the entire platform
could collapse. The ether’s exchange rate was in freefall. Buterin undoubt-
edly felt that something had to be done.
Yet it wasn’t obvious that anything could be done. After all, The DAO
was built from Ethereum’s “immutable, unstoppable, and irrefutable
computer code.” The entire point of the platform was that there wasn’t
any admin panel that someone could call up to cancel transactions that
they didn’t like. Contracts were executed as written. Those were the rules.
And it was next to impossible for even Buterin to break Ethereum’s rules.
However, changing the rules was a different matter. Small changes to
the rules were being made almost routinely as part of software updates
that Buterin and his team issued to users. It was possible to imagine a more
complex rule change that would in effect reverse the effects of the hack by
forcing the misappropriated funds to return to the original investors— like
a special law that said, “The contract signed on this date between these par-
ties shall be deemed null and void, and any funds transferred thereunder
shall be returned to their original owners.” It would be a complex software
update, but the process of implementing it would be the same as with ear-
lier updates: Buterin’s team releases an update, users download and install
it on their computers, and the new rules take effect. From that point on,
the decentralized platform would enforce the new rules with the same
“steadfast iron will” with which it had enforced the previous rules up to
that point. Problem solved.
This is a portion of the eBook at doi:10.7551/mitpress/14219.001.0001
Downloaded from http://direct.mit.edu/books/book/chapter-pdf/2423278/c005400_9780262371094.pdf by CSU: San Francisco user on 26 August 2024
144 CHAPteR 8
Of course, the slight issue was that this update would reveal the whole
idea of immutable records and unstoppable code as an illusion. For what
good were rules administered without any possibility of human interven-
tion if humans could change those rules at will?
The update was thus likely to cause some contention. And to Buterin’s
team’s credit, they did not attempt to simply impose it on Ethereum’s
users. Instead, they began to publicize the problem and their proposed
solution via discussion forums and blog posts. Thanks to a safeguard built
into The DAO’s code, the hacker wouldn’t be able to spend their crypto-
takings for another four weeks, so there was still time to deliberate.
Many users agreed with the proposed rule change. Others opposed it on
the grounds that changing the rules retroactively seemed to run counter to
the platform’s whole idea. In the end, Buterin organized an ad- hoc online
referendum. Users’ voting power was proportional to how much of the
ether cryptocurrency they owned. Votes representing only about 6 percent
of all ether in circulation were cast. Despite the publicity, it is likely that
less active ether owners didn’t hear about the hastily organized vote. In any
case, the yeas beat the nays almost seven to one, and the update was car-
ried out. The DAO’s funds were returned to their original owners.
The crisis revealed how a peer- to- peer blockchain system in the end
was never really “trustless.” The network may have enforced its rules with
robotic impartiality, but people were still in charge of making and amend-
ing the rules. In this instance, people decided to amend the rules to con-
fiscate a person’s holdings and return them to their previous owners. The
point here is not whether this decision was justifiable or carried out in
a democratic fashion. The point is that it was possible in the first place.
Funds placed in the system were still ultimately entrusted to the care of
people, not cryptography. The problem of trust remained unsolved.
Ethereum did survive but with a bruised reputation. Much soul- searching
ensued. The word “unstoppable” was removed from the platform’s home
page. “It turns out we have a lot in common with central banks,” com-
mented a former Ethereum project manager. “Maybe not at the technical
or legal level, but at a political level, people in our community expect us
to be able to make things better for them.”38
Similar incidents showed how Bitcoin likewise still ultimately depended
on human rulemaking. When Nakamoto withdrew from the project, he
This is a portion of the eBook at doi:10.7551/mitpress/14219.001.0001
Downloaded from http://direct.mit.edu/books/book/chapter-pdf/2423278/c005400_9780262371094.pdf by CSU: San Francisco user on 26 August 2024
CRyPtoCRACy 145
handed it over to Australian American programmer Gavin Andresen.
Andresen appointed a team of software developers to help him. This core
team issued software updates that added new features and fixed bugs.
Andresen downplayed the team’s role, suggesting that they merely took
care of the “plumbing.”39 But one bug was so severe that it would have
allowed an attacker to generate new bitcoins at will.40 In cases like this, the
team didn’t necessarily tell the users the full story of the update’s purpose
beforehand. Users were in effect asked to trust the team. Bitcoin, too, was
ultimately not “math- based money” but people- based money, not fun-
damentally dissimilar from the pounds and dollars that Nakamoto had
sought to replace.
COMPETING INTERESTS
Around the same time, it was starting to become apparent that even
blockchain’s revolutionary rule- enforcement system wasn’t quite as trust-
less as Nakamoto had thought.
The idea of using a proof- of- work scheme to select the administrator in
charge of recordkeeping seemed like a brilliant bit of engineering. It meant
that selection was random but eligibility was tied to something tangible—
CPU power. As in ancient Athens, participation was incentivized with
rewards: fresh bitcoins were awarded to the successful appointee each time
the selection was performed. Nakamoto called this process mining:
The steady addition of a constant amount of new coins is analogous to gold miners expending resources to add gold to circulation. In our case, it is CPU time and electricity that is expended.41
Miners’ rewards could be substantial. In 2015, the reward per block
was twenty- five bitcoins, worth over $6,000 on the trading sites. Since a
new block was mined once every ten minutes, the total payouts added up
to over $6 million per week. Mining quickly began to attract professional
interest. Data center– style industrial cryptomining outfits emerged, kit-
ted with custom- built hardware and bulk electricity access.
However brilliant a cryptographer Nakamoto was, he clearly was not
an economist, for he did not realize that just like gold mining, bitcoin
mining would entail economies of scale.42 Industrial mining operations
incurred much lower unit costs than individual users with ordinary PCs
This is a portion of the eBook at doi:10.7551/mitpress/14219.001.0001
Downloaded from http://direct.mit.edu/books/book/chapter-pdf/2423278/c005400_9780262371094.pdf by CSU: San Francisco user on 26 August 2024
146 CHAPteR 8
did. The industrialists thus quickly outcompeted the ordinary users whom
Nakamoto had expected to shoulder the system’s administration. Instead
of circulating randomly between thousands of cryptocitizens, Bitcoin’s
official recordkeeping duties began to cycle between a handful of large
corporations.
At the end of 2015, just three companies were responsible for mining
60 percent of Bitcoin’s record blocks. In principle, anyone controlling over
half of the mining power would have been able to stop any and all transac-
tions they didn’t like, holding the network hostage.43 Managers who repre-
sented approximately 90 percent of the network’s mining power appeared
on a stage together at a Bitcoin conference in Hong Kong in December
2015. The managers sought to assure the citizens that they had the net-
work’s best interests in mind. “Trustless” recordkeeping had turned into
“trust us.”
Bitcoin’s lead developer Andresen argued that the concentration of min-
ing power into the hands of a few large corporations was not a big deal
because it would not make economic sense for a mining company to under-
mine the system from which its profits derived.44 Still, it meant that if some-
thing should ever emerge to threaten the companies’ profits, they would
not be powerless. And it turned out that Andresen himself was about to
trigger such a scenario and end up feeling the mining corporations’ power.
In 2015, Andresen proposed increasing the size of the Bitcoin blocks
on which transactions were recorded. The rationale was simple. At that
time, each block could accommodate at most about two thousand trans-
actions. Given that the system was designed to add a new block to the
chain once every ten minutes or so, this meant that the Bitcoin network
was able to confirm only about 3.5 transactions per second. This had
been more than enough at first, but now the network’s increasing popu-
larity had led to congestion. People sometimes had to wait hours for their
payments to be confirmed, making the system practically unusable.
Andresen proposed increasing the maximum block size twentyfold,
resulting in a capacity of about seventy transactions per second. This would
still be a Lilliputian capacity compared to a mainstream payment system
like Visa, which processed two thousand transactions per second on aver-
age and had a maximum capacity of 56,000 transactions per second. But
This is a portion of the eBook at doi:10.7551/mitpress/14219.001.0001
Downloaded from http://direct.mit.edu/books/book/chapter-pdf/2423278/c005400_9780262371094.pdf by CSU: San Francisco user on 26 August 2024
CRyPtoCRACy 147
it would be a simple update to implement and would at least alleviate the
problem for the time being, Andresen argued.
It turned out that a powerful interest group among Bitcoin’s stake-
holders was against such a change. Thanks to the availability of cheap
government- subsidized electricity, the mining industry had become heav-
ily concentrated in China. At the end of 2015, about three- quarters of all
the mining power in the Bitcoin network originated from China.45 The
country’s Great Firewall restricted Internet bandwidth between Chinese
mining companies and the rest of the world. This meant that Andresen’s
proposed larger blocks would have been difficult for Chinese miners to
handle. “An increase in block size to 20 megabytes would increase operat-
ing costs for miners,” explained one Chinese mining executive.46
Moreover, mining companies everywhere actually profited from the
network’s congestion, at least in the short term. Miners had the power to
choose which transactions from the queue of pending transactions they
included in the blocks that they produced. Ordinary users, desperate to get
their transactions picked up ahead of the queue, furnished their payment
instructions with “tips” that miners could collect when they processed
the payment. The worse the congestion became, the bigger the tips users
were willing to offer to bypass it. Major mining corporations quickly sided
against Andresen’s proposal. Andresen arranged talks with them but to
no avail.
A variety of other commercial and ideological interests also hinged on
the block size question. Some interest groups publicly expressed support
for Andresen’s proposal. Others opposed it. But Bitcoin had no formal
decision- making processes— that is, formal political institutions— that all
sides would have considered legitimate and thus no way of reconciling
the conflict. Divisions intensified. Rhetoric hardened. Debate broke down
into tribalism, trolling, and social media bot campaigning. A climax of a
sort was eventually reached when another developer betrayed Andresen’s
trust by canceling his write access to Bitcoin’s official code repository,
effectively throwing him out of the core team.
A Chinese mining company executive lamented:
A decentralized system . . . needs a democratic mechanism to operate and to avoid that disputes are thrown into the Bitcoin community directly and rudely.47
This is a portion of the eBook at doi:10.7551/mitpress/14219.001.0001
Downloaded from http://direct.mit.edu/books/book/chapter-pdf/2423278/c005400_9780262371094.pdf by CSU: San Francisco user on 26 August 2024
148 CHAPteR 8
A BROKEN MARKET FOR RULES
Decision making in open- source software development projects is often a
mix of two contrasting elements. One element is strong technocratic direc-
tion by a skilled and charismatic lead programmer, sometimes referred to
as a benevolent dictator.48 For instance, Vitalik Buterin is widely recognized
as Ethereum’s benevolent dictator.49 The dictator and their team usually
control the project’s infrastructure, such as its official communication
channels and code repository. This concentration of power can provide
efficient decision making toward a consistent vision.
The other decision- making element is so- called rough consensus— an
informal norm that any significant changes to the software should enjoy
near- unanimous support from the “community”. The community is never
clearly defined, but it can mean software developers actively working on
the project, companies using the software, and sometimes even individ-
ual users, depending on whom you ask. This popular assembly of a sort
acts as a check on the executive power and helps to ensure that decisions
are informed by a broad range of perspectives.
In practice, the popular assembly usually takes the form of deliberation
via a mailing list or an online forum. The goal of the deliberation is to reach
a rough consensus on any major issues at hand, on which basis the dicta-
tor may then act. Rough consensus means that little or no disagreement
remains among those participating in the debate. Majoritarian decision
making of the sort used in modern democracies— where a vote is held and
the minority must accept the view of the majority— is not as frequently
used, partly because it is unclear who should be eligible to vote. The fol-
lowing motto, coined by influential Internet engineer David D. Clark— and
repeated by many blockchain developers— expresses this ideal:
We reject: kings, presidents, and voting. We believe in: rough consensus and running code.50
This combination of two contrasting elements— or in social science
terms, contrasting political institutions— has evolved through decades of
open- source software development and clearly it presents some advan-
tages. But it remains poor at reconciling conflicts. No matter how frus-
trated people get with benevolent dictators, there is no process for
replacing them. This has allowed some “benevolent” open- source dictators
This is a portion of the eBook at doi:10.7551/mitpress/14219.001.0001
Downloaded from http://direct.mit.edu/books/book/chapter-pdf/2423278/c005400_9780262371094.pdf by CSU: San Francisco user on 26 August 2024
CRyPtoCRACy 149
to bully community members for years (fortunately this is not alleged in
Bitcoin’s case or in Ethereum’s case). And insofar as community decision
making relies on consensus and rejects voting, in contentious issues it eas-
ily becomes deadlocked.
However, open- source software projects carry a third political institution
in their back pocket, which can be pulled out in case of irreconcilable con-
flict: the fork. Since all source code is freely available, developers who are
not happy with the leadership or direction of the venture can simply copy
the code and launch their own version. The two parallel projects diverge
from their common haft, like tines on an eating implement. Because of
the possibility of forking, open- source software projects are sometimes
thought of as anarchic or even democratic even as they are being overseen
by dictators.
In the same way, blockchain proponents sometimes maintain that their
projects are “trustless,” even as they rely on powerful lead developers and
mining corporations, because anyone unhappy with the arrangement
could in principle launch their own fork. Indeed, this is what Gavin Andre-
sen’s allies did when they lost the fight over Bitcoin’s block size: they cre-
ated their own parallel version of the Bitcoin software with different rules
and set up a parallel peer- to- peer network with that software. Similarly,
people who disagreed with Vitalik Buterin on changing Ethereum’s rules
to countermand The DAO contract set up their own Ethereum network
where the update never happened. People could then choose which of the
two parallel Bitcoins or Ethereums they preferred— and thus which set of
rules they were subject to. In line with anarcho- capitalist thinking, “the
market” would choose the rules; there was no need for formal political
institutions like voting.51
The market for rules was not very liquid, however. One problem was
that the platforms’ value was based on network effects.52 The most useful
payment platform for consumers was the one that businesses used and
vice versa. In other words, individuals couldn’t simply choose the systems
that they personally preferred but had to take others’ choices into account
also. When the two Bitcoin networks split from each other, miners and
users at first wavered between the two alternatives (figure 8.1).53 But as it
started to become clear which side was going to emerge as the de facto stan-
dard, the overwhelming majority quickly gravitated to that one, and the
This is a portion of the eBook at doi:10.7551/mitpress/14219.001.0001
Downloaded from http://direct.mit.edu/books/book/chapter-pdf/2423278/c005400_9780262371094.pdf by CSU: San Francisco user on 26 August 2024
150 CHAPteR 8
competing “coin” lost almost all of its support and value. The same hap-
pened in the Ethereum split. Individuals could not freely choose which
rule set to follow; in the end they had to follow the majority if they were
to continue transacting. Choice between institutions is not an individual
choice but a collective one.
But must a choice be made? Couldn’t people simply use multiple sys-
tems in parallel? This is called multihoming in platform theory. In some
contexts, multihoming is viable, and it helps to limit the winner- takes- all
dynamics that result from positive network effects. Using multiple pay-
ment systems in parallel is clunky but not impossible.
But when a blockchain system is supposed to record ownership stakes
in assets like land titles, stocks, or nonfungible tokens (NFTs), multihom-
ing after a fork becomes untenable. Suppose that The DAO had already
bought some stocks before the hack happened and the Ethereum net-
work split into two. Now there are two duplicate versions of the Ethereum
blockchain, both purporting to contain records attesting to ownership of
the same stocks. Suppose further that people multihome and both chains
remain in operation. On one chain, the owners of a stock sell it. On the
other chain, they don’t sell it. The same stock now has different owners
Bitcoin cash
Network splits
2018 2019 20202017
S h
ar e
o f
m in
in g
p o
w er
Bitcoin core
BitcoinBitcoinBitcoin
8.1 Bitcoin cash versus Bitcoin core share of total mining power, 2017 to 2019.
This is a portion of the eBook at doi:10.7551/mitpress/14219.001.0001
Downloaded from http://direct.mit.edu/books/book/chapter-pdf/2423278/c005400_9780262371094.pdf by CSU: San Francisco user on 26 August 2024
CRyPtoCRACy 151
in the two different blockchains. Which record is correct? The technology
provides no answer. People must somehow choose one chain to be the
authoritative one and discard the other.
Still, even if we accept that a choice must be made and that the choice
between institutions is a collective one, then blockchain network splits
could be seen as referenda of a sort on which rules the collective would
like to adopt. The only eligible voters in such referenda are the miners,
though, and indeed the only thing that Nakamoto’s original Bitcoin paper
said about the system’s governance was that miners would “vote with
their CPU power.”54 Ordinary users can influence the referendum result
only through informal and indirect means, such as by buying a particular
token to try to push up its value to incentivize miners to mine it.
Moreover, nothing in the technology guarantees that users get to make
an informed choice. When Buterin’s team at the Ethereum Foundation
created a new version of the software that changed the rules of the game,
the foundation’s trademark ensured that only this new software would be
called Ethereum. The version that remained unchanged had to adopt a
new name, even though it was the one that represented continuity. In the
case of Bitcoin, some community moderators attempted to prevent users
from knowing that there was any choice to be made at all by banning all
discussion of Andresen’s allies’ version.
In both conflicts, the eventual winner was the group in control of the
things that could not be split, such as the official code repository, the offi-
cial communication channels, and— crucially— the system’s official name
and branding. The trading sites also played a kingmaker role by decid-
ing which version gets to keep the established ticker symbol and which
one must adopt a new one. For all these reasons, forking does not make
blockchain networks “trustless,” nor is it an effective substitute to formal
political institutions. The Bitcoin split, in particular, played out more like
a civil war than a referendum; naked power was deployed at least as much
as argument.
TRUSTED CENTRAL PARTIES
Bitcoin’s capacity to handle transactions remains extremely limited to this
day. It was never widely adopted as a payment system. Many shops and
This is a portion of the eBook at doi:10.7551/mitpress/14219.001.0001
Downloaded from http://direct.mit.edu/books/book/chapter-pdf/2423278/c005400_9780262371094.pdf by CSU: San Francisco user on 26 August 2024
152 CHAPteR 8
restaurants that once experimented with the currency stopped accepting
it.55 Staff at the vegan café near where I lived told me that it was mostly
the journalists writing stories about Bitcoin who paid with it in the first
place. Instead of being used for payments, Bitcoin and other cryptocurren-
cies turned into speculative investment assets. People buy bitcoins in the
hopes that someone else would later buy the coins from them for an even
higher price.
The great majority of cryptocurrency investors today don’t actually hold
the keys to the coins that they buy. Like stuffing cash in a mattress, holding
cryptocurrency is risky and inconvenient. Instead, they entrust their coins
to a handful of companies that run the largest trading sites and access the
funds by logging into the equivalent of online banking. The vast majority
of cryptocurrency transactions happen inside the proprietary systems of
these new financial institutions. The Bitcoin network functions not as a
payment system for ordinary people but as a sort of interbank settlement
network between these institutions.
The leading trading sites and mining companies— many of them owned
by the same people— now measure their profits in billions of dollars. The
crypto- elite who run these organizations are, if anything, less account-
able to the people than conventional financial and regulatory elites. They
are caught lying to their customers, defrauding them, manipulating the
market, and peddling assets they know are not backed by sufficient collat-
eral, and yet the show goes on.56 In the eventual crisis, millions of people
will again lose some or all of their savings, while insiders’ profits will have
been long since off- shored. The Electronic Frontier Foundation, true to
its cyberlibertarian form, lobbies against government intervention.57 The
only saving grace is that for now cryptomarkets remain small compared
to mainstream financial markets, limiting the damage.
None of this to suggest that Bitcoin’s or Ethereum’s creators themselves
were untrustworthy or insincere people. Compared to platform company
barons like Amazon’s Jeff Bezos or PayPal’s Peter Thiel, they were remark-
ably open about their plans, gave many users a voice in decision making,
and solicited user consent for many important decisions. The point is sim-
ply that like previous cypherpunks and crypto- anarchists, they ultimately
failed to eliminate trust as something that underpins economic activity,
with unfortunate consequences.
This is a portion of the eBook at doi:10.7551/mitpress/14219.001.0001
Downloaded from http://direct.mit.edu/books/book/chapter-pdf/2423278/c005400_9780262371094.pdf by CSU: San Francisco user on 26 August 2024
CRyPtoCRACy 153
Meanwhile, another unintended consequence of Nakamoto’s proof- of-
work scheme has become impossible to ignore. Mining corporations’ total
electricity consumption now rivals the electricity needs of medium- sized
countries.58 The proof- of- work algorithm causes a network’s energy con-
sumption to be proportional to its coin’s exchange rate. As long as crypto-
investors keep paying fabulous sums for the coins that miners mint, mining
corporations will keep burning energy like there’s no tomorrow. More effi-
cient mining hardware simply results in larger amounts of hardware being
used to burn the same amount of energy. “Proof-of-work” should really be
called “proof-of-waste.” China eventually banned cryptomining, but the
industry moved into other countries, led by the United States.59 Ethereum’s
developers have been trying to implement a less damaging scheme for
years. Bitcoin’s developers have not announced any such plans. Promi-
nent Bitcoin proponents focus on downplaying and denying the system’s
climate implications.
RISE OF CRYPTOCRACY
Satoshi Nakamoto wanted to solve the problem of trust by delegating
power to an incorruptible machine— a peer- to- peer blockchain network.
Like the kleroterion of ancient Athens, the machine would distribute
administrative responsibilities to so many people that none individually
would wield power sufficient to coerce another, creating order without risk
of abuse. This decentralization of administration didn’t quite succeed, in
that economies of scale concentrated power into the hands of a few large
companies. But even if it had succeeded, it still would not have eliminated
trusted authorities from the system, because Nakamoto did not design
anything comparable to the other crucial aspect of Athenian democracy:
decentralization of legislation.
Legislation and administration are two sides of a coin. Legislation cre-
ates rules, while administration applies them. Solon designed institu-
tions for both. Nakamoto was so concerned with creating an incorruptible
administration that he paid no attention to legislation. While adminis-
tration can to some extent be automated, legislation cannot. “You will not
find a solution to political problems in cryptography,” somebody on the
mailing list warned Nakamoto when he first announced his project.60 But
This is a portion of the eBook at doi:10.7551/mitpress/14219.001.0001
Downloaded from http://direct.mit.edu/books/book/chapter-pdf/2423278/c005400_9780262371094.pdf by CSU: San Francisco user on 26 August 2024
154 CHAPteR 8
Nakamoto missed the point and proceeded as if politics in his system
didn’t exist.
For many years, then, blockchain developers believed— or at least
pretended— that they were mere plumbers. To acknowledge that they were
in fact politicians— on whose decisions so many fortunes now hung— ran
counter to their crypto- anarchist creed. Thus, they tried to keep everything
informal and resisted creating formal political institutions that could have
distributed legislative power more widely but would have also revealed
their own de facto power.
Crises like The DAO attack and the block- size conflict finally forced
developers to confront their blockchains’ politics. They tried to gov-
ern their software as a traditional open- source project, but into the mix
now entered investors, trading sites, billion- dollar mining corporations,
shady Bahamian banks, and other stakeholders with financial interests
and resources to spend. Open- source dissidents’ traditional weapon— the
fork— was dulled in the presence of network effects, and it was completely
off the table for any blockchain intended to maintain a definitive record
of who owned what in the real world. In the absence of formal processes,
many important decisions turned into backroom politics and social
media warfare. Most ordinary users had no idea who the systems’ power
blocs were, what goals they pursued, or whose social media accounts they
funded. Most journalists continued to write stories of math- based money
that somehow ran itself.
In his quest to eliminate trusted authorities, Nakamoto succeeded
mainly in obscuring who the authorities were. His complicated attempt
at substituting technological certainty for human fallibility resulted in
such a convoluted system that power holders became difficult even to
recognize, let alone call to account. His pseudonym and Solon- like van-
ishing act conjured a legend over his creation, which further obscured
its workings. In attempting to forgo the need for popular rule as in the
Athenian dēmokratía, he instead ended up enabling a regime of secretive
rule— a kryptókratía.
This is a portion of the eBook at doi:10.7551/mitpress/14219.001.0001
Downloaded from http://direct.mit.edu/books/book/chapter-pdf/2423278/c005400_9780262371094.pdf by CSU: San Francisco user on 26 August 2024
This is a section of doi:10.7551/mitpress/14219.001.0001
Cloud Empires How Digital Platforms Are Overtaking the State and How We Can Regain Control
By: Vili Lehdonvirta
Citation: Cloud Empires: How Digital Platforms Are Overtaking the State and How We Can Regain Control By: DOI: ISBN (electronic): Publisher: Published:
Vili Lehdonvirta
The MIT Press 2024
10.7551/mitpress/14219.001.0001 9780262371094
Downloaded from http://direct.mit.edu/books/book/chapter-pdf/2423278/c005400_9780262371094.pdf by CSU: San Francisco user on 26 August 2024
© 2022 Vili Lehdonvirta
All rights reserved. No part of this book may be reproduced in any form by any
electronic or mechanical means (including photocopying, recording, or information
storage and retrieval) without permission in writing from the publisher.
The MIT Press would like to thank the anonymous peer reviewers who provided
comments on drafts of this book. The generous work of academic experts is essential
for establishing the authority and quality of our publications. We acknowledge with
gratitude the contributions of these otherwise uncredited readers.
This book was set in Stone Serif and Avenir by Westchester Publishing Ser vices.
Library of Congress Cataloging-in-Publication Data
Names: Lehdonvirta, Vili, author.
Title: Cloud empires : how digital platforms are overtaking the state and
how we can regain control / Vili Lehdonvirta.
Description: Cambridge, Massachusetts : The MIT Press, [2022] | Includes
bibliographical references and index.
Identifiers: LCCN 2021058948 | ISBN 9780262047227 (hardcover)
Subjects: LCSH: Cyberspace—Social aspects. | Cyberspace—Economic aspects. |
Digital media. | Central planning. | Power (Social sciences)
Classification: LCC HM851 .L444 2022 | DDC 303.48/34—dc23/eng/20220421
LC record available at https://lccn.loc.gov/2021058948
MIT Press Direct
Downloaded from http://direct.mit.edu/books/book/chapter-pdf/2423278/c005400_9780262371094.pdf by CSU: San Francisco user on 26 August 2024
