Root of Trust
By Justin Maier on ALTCOIN MAGAZINE
Blockchain exists to create trusted records. To ensure that a trustworthy record is maintained, blockchains employ a few different methods. To help illustrate how each of these work, I’m going to use the analogy of playing a board game.
Nothing
Technically
We trust each other to make the transaction
Board game
We read the rules and trust one another to play fair. If someone cheats and we notice, we call them out and the game is “done” or we try to undo the cheating.
Third-party
Technically
We trust a third-party to make the transaction between us.
Board game
We read the rules and to make sure that we follow the rules, we get another person to read the rules and then make the moves for us. If someone cheats and the person notices, they prevent the move.
Blockchain: Proof-of-Work
Technically
A pool of machines watches every transaction and ensures that the transaction can be done and provides mathematical proof. Every machine in the network reaches consensus and the chain continues.
Board game
We read the rules and have a stadium of 10 million people getting paid to watch us. Every time we make a move, each person in the stadium checks to see if that move was valid. If one of us cheats, the people in the stadium will undo the cheating move.
Blockchain: Proof-of-Stake
Technically
Operates technically similar to Proof-of-Work, except to expedite the process, the owners of the machines doing the proofs must “stake” (put forward money) to prove that they will behave. If they misbehave, they lost what they staked.
Board game
We read the rules and have 1,000 paid guards that say, “if I let anyone cheat you can take this money I brought with me”. If one of us cheats, the guard will undo the cheating move.
Blockchain: Proof-of-Authority
Technically
Operates technically similar to Proof-of-Work, except to expedite the process, only authorized machines can do the proofs.
Board game
We read the rules and have someone that the game-maker selects watch us. If one of us cheats, then they undo the cheating move.
After walking through the scenarios, I hope you notice a few things:
- We have to get someone else involved if we don’t trust one another.
- The amount of trust between one another increases as the number of people involved in verifying that trust increases.
- The more people that are involved in verifying trust, the more inefficient it is to interact with one another.
- Only with a trusted third-party are the “bad” moves prevented before they happen.
There is a balance between trust and efficiency that finds equilibrium in a single third-party. By having a single person ensure that our moves abide by the rules, we are able to play together quickly without having to trust that the other person is following the rules. But what if we don’t trust the third-party? We could add more people to verify the moves, but in doing so it increases the time it takes to verify the moves and play the game. This reality forces us to exchange efficiency for trust in an increasingly less trust-worthy world. But what if we could ensure that someone was always trustworthy?
We as humans are not and never will be always trustworthy, however, we have created machines that can be. There are several ways to make a trustworthy machine. With blockchain, we ensure that the machines are trustworthy by requiring them to reach consensus. However, all of the methods used by blockchain either give up trust for efficiency or vice-versa. Development continues to go into improving the efficiency of blockchain, but what if there is a better way?
Cryptography was created as a way to securely interact. One of the simplest forms of cryptography is creating ciphers or “keys” that are used to encrypt messages. With symmetric encryption, we can know or trust that only someone with the key can read or write those messages. Using those two points of trust, we can create a trustworthy machine.
A cryptographic trust machine is pretty simple. When someone sends a message, it encrypts it using a key that only it holds and stores it for later reference. In the future, if someone doubts that message happened, it can try to decrypt the message. If it decrypts successfully, then we can know that it was a valid message and it can confirm that it happened. Since only it can create and read the records, the unique key becomes a root of trust. Once the root of trust exists, the key can evolve to retain its security and become more difficult to crack.
To understand how a cryptographic trust machine works, let’s go back to our board game example. We read the rules and to make sure that we follow the rules, we have a machine that has been programmed with the rules make the moves for us. To make sure no one has cheated, each valid move receives a stamp of approval that only the machine can give. If someone wants to cheat, the machine will see that it’s against the rules and won’t make the move. If someone tries to manually move, it will be obvious because it doesn’t have the stamp of approval.
With the cryptographic trust machine acting as an intermediary, we can reach the perfect balance of trust and efficiency in every interaction. By programming a set of rules into the trust machine, we can ensure that only records that abide by those rules will be added to the machine. Those rules could be anything including rules about how to change the rules.
