Cryptography scales trust
Protocols are to institutions as packet switching is to circuit switching
People who work in cryptography often talk about zero-trust protocols, but what they mean is: cryptography scales trust.
How? Practically speaking, cryptography lets us solve a bunch of thorny problems with trust, stuff like:
Authenticity: how do I know you are who you say you are? Cryptographic signatures use math to prove a message was signed with a particular cryptographic key. Signing makes messages non-repudiable, meaning the author is provably the author. This holds true even if the message was sent from a computer or network that can’t be trusted.
Privacy: how do I know only you can see it? End-to-end encryption (e2ee) encrypts data so that only the recipient who holds the key can decrypt it. To everyone else, it’s a black box that can’t be read or tampered with.
Integrity: how do I know this is it what it says it is? Is this mp3 a song or a virus? Cryptographic hashing functions let us generate a hash, a string of numbers and letters, unique to the content. To verify the integrity of the file, we just check the hash. If the file has been tampered with, the hash won’t match.
Proof-of-x: how do you show you’re human? Authorized to log on? Eligible to vote? ZK proofs let people cryptographically prove things about themselves without revealing their actual identity or information.
Contracts: how do I know you’ll do what you say you’ll do? Smart contracts deterministically execute agreements when conditions are met, preventing defection or cheating.
These are just some examples. At the most abstract level, cryptography allows us to replace other forms of trust with trust in math. Or put another way,
Cryptography is a tool for turning lots of different problems into key management problems.
(Dr. Lea Kissner, Head of Privacy Eng and CISO at Twitter)
We’ll get to that in a minute. Anyway, the trust isn’t eliminated. It just moves to a different part of the system, a part which can be scaled.
Take your bank website, for example. How do you know it’s legit? After all, the website had to hop through a lot of computers to get to you. What if one of them swapped out the real website for something else? A scam? Well, instead of trusting each computer that helped deliver the website, you trust the cryptographic key that was used to sign it. That little green lock in your browser? That’s called SSL.
It’s not clear how we could even establish trust on the internet without cryptography. To trust a message, I would have to trust every computer that touched the message as it bounced around the internet to me. Since internet routing is dynamic, I can’t even know who that might be ahead of time!
So, ordinary forms of trust, like reputational trust, just taking your word for it, cannot work at internet scale. But trusting a cryptographic key? That works! Cryptography scales trust.
Clever hacks to scale trust
Finding new ways to scale trust is a big deal. In a state of nature, our species lives in small bands of 150 or less, hunting, gathering, sharing stories, migrating from one place to another. Yet today, it is common to live in cities of a million or more. How did we end up here?
Well, one way to see our history is as a series of clever hacks to scale trust. Each unlocked new forms of cooperation, leading to new social structures. Researcher David Ronfeldt identifies at least four:
How have people organized their societies across the ages? The answer may be reduced to four basic forms of organization,
the kinship-based tribe, as denoted by the structure of extended families, clans, and other lineage systems.
the hierarchical institution, as exemplified by the army, the (Catholic) church, and ultimately the bureaucratic state.
the competitive-exchange market, as symbolized by merchants and traders responding to forces of supply and demand.
and the collaborative network, as found today in the web-like ties among some NGOs devoted to social advocacy
(Ronfeldt, 1996, “TIMN - Tribes, Institutions, Markets, Networks”)
We’re still figuring out that fourth one. Anyway, each of these social structures was catalyzed by a hack to scale trust:
Language scaled trust by enabling oral storytelling, the recitation of genealogies, and the development of larger kinship networks. I can trust you because my father’s father’s father is your father.
One of the fascinating characteristics of lineages is that they can be aggregated upward into much larger superlineages simply by tracing descent back to an earlier ancestor. (Fukuyama, 2011. The Origins of Political Order)
Shared language lead to the accumulation of shared ideas, shared beliefs, shared gods. Belief in Big Gods scaled trust by expanding the scope of kin to include co-religionists. I can trust you, because we are brothers and sisters in the same god and share the same moral framework. Ideologies perform the same purpose.
The ability to create mental models and to attribute causality to invisible abstractions is in turn the basis for the emergence of religion. (Fukuyama, 2011)
If we had to constantly negotiate new rules with our fellow human beings at every turn, we would be paralyzed and unable to achieve routine collective action. The fact that we become attached to certain rules not as means to short-term goals but as ends in themselves greatly enhances the stability of social life. Religion simply reinforces that stability and widens the circle of potential cooperators.
(Fukuyama, 2011)
Writing scaled trust by expanding the reach of Big Gods, and by enabling the emergence of durable hierarchical institutions. I can trust this because it was written down. It isn’t just hearsay. It is the word of god, etched in stone, or perhaps the word of an institutional authority, inked in papyrus, signed with an official seal.
Writing also meant math, and math lead to money. Money scaled trust by creating markets. I don’t have to trust you, I can trust double-entry bookkeeping and the almighty dollar.
These are just a few of the tricks for scaling trust that allowed us to solve new and bigger coordination problems. More trust meant greater levels of cooperation, deeper specialization, more social complexity, larger economies of scale.
Of course, this specialization and hierarchy and complexity has its downsides. Sometimes modernity makes me want to return to monke. Yet, it is difficult to outrun the economies of scale that come through scaling trust.
So, tribes, institutions, markets. What about networks? I want to propose that,
Cryptography scales trust by making it possible to build networks that only require trust at the the endpoints.
When the network is built on cryptographic protocols, we don’t have to trust anything in-between. We don’t have to trust the computers, the wires, the institutions, beliefs, or kinship of the network participants. We can trust the math. All we have to do is verify the message with cryptography.
This is a powerful primitive for scaling trust, because it means everyone on the network can cooperate together, even if they don’t trust each-other. The costs to producing high-trust results are greatly reduced, and we can spend our coordination efforts on building larger, more fine-grained, more complex networks of cooperation.
If language produced tribes, writing built institutions, money made markets, then perhaps cryptographic protocols will construct a network society?
The crypto archipelago
In a high-trust society, we might not often care about cryptographic. In such a society, social protocols can take the place of cryptographic ones, and we can scale trust through means other than math: I trust in God, I trust Google, I trust government, I trust judges to rule with an even hand, I trust contracts, I trust institutions, I trust banks, and money I’m holding in my hand.
Nice work if you can get it. High social trust creates something like a Coasean “firm”, within which the costs of cooperation are greatly reduced. The higher the trust, the larger the Coasean space in which we can cheaply cooperate.
Does this look like a high-trust society?
The challenge is that we seem to be in the middle of some kind of social transition, driven by the internet. Old forms of trust are eroding, or getting DDOS’d by the scale of the internet. Social transitions are messy.
A new internet-scale paradigm, the cryptographic protocol, is in its nascent stages. But then new paradigms, by their very nature, question the logic of incumbent social structures which were designed around older forms of trust.
All media work us over completely. They are so pervasive in their personal, political, economic, aesthetic, psychological, moral, ethical, and social consequences that they leave no part of us untouched, unaffected, unaltered.
(Marshall McLuhan, 1967. The Medium is the Massage)
The closest analog to this transition might be the introduction of the printing press in Europe. This transformed a world of feudal estates, knit together by religion, into a world of states, knit together by markets, communication networks, and scientific/democratic institutions. I like these changes a lot! But then they also induced 200 years of religious warfare. Yikes.
So it might be the case that we find ourselves in a high-trust society that is experiencing rapid unbundling. Formerly functional institutions and shared social norms are fragmenting into pieces.
When hierarchies break down, they usually split along their subsystem boundaries. (Donella Meadows, Thinking in Systems)
What we’re left with is the small remaining islands of functional high-trust—friends, family, maybe church, community, some local institutions—separated by growing oceans of distrust. Where to go from here?
Cryptography suggests a way forward: network.
Cryptographic protocols can act like undersea cables between islands of high-trust. We can span low-trust oceans because we only have to trust the endpoints. Trustless networks re-build trust on a new foundation, giving it new scaling properties. You don’t have to trust apps not to snoop, you can trust the end-to-end encryption. You don’t have to trust me to honor the contract, you can trust the smart contract code. Parties with limited trust between can cooperate effectively together using protocols. The protocols knit them together into a functional network organization.
We could think of cryptographic protocols as a kind of software-defined Weberian routinization, where protocols are to institutions as packet switching is to circuit switching.
Queries
I’m surrounded by smart readers, and so I would love your help in answering some of the questions I’ve been exploring:
Who are the smartest people working in this space? Who is building new networked forms of coordination? Bonus for examples with fundamental needs in the loop—food, water, shelter, etc. Double bonus for street uses, as in “The street finds its own uses for things” (William Gibson)
“A good science fiction story should be able to predict not the automobile, but the traffic jam." (Frederik Pohl) What traffic jams should we anticipate in the transition toward network society? Where could cryptographic protocols help? What traffic jams will protocols cause?
What are some other ways we’ve scaled trust? What was the transition like? Who’s studied this? What should I be reading?
Reply to this email and let me know!