From trusting people to trusting processes

The emergence of the concept of the blockchain, that was first introduced as a decentralized ledger for the bitcoin currency, leads us to think about the different models upon which trust is built in organizations. I think that all kinds of organizational trust could be reduced to 3 main models:

1.  Good-Heartedness-Based Trust - This is when we trust that someone would not harm us, because this person is a goodfella who would never do that to us, not willingly at least. We could also call this the "History-based model", because it is based on the history of that person; someone with an honest history will probably continue to be honest.
2. Disinterest-Based Trust - We trust that someone will not harm us if it will also bring harm to them. For example, a company would not sell me a defected product, if they know this will damage their reputation.
3. Process-Based Trust - We trust that no one can harm us, because we are protected by a process that disallows them to do so. For example, if I have bodyguards, I am protected by a process of big people beating up other people who try to approach me. Another example is the Parliament, which is a process to protect against a single person making decisions that will harm others. A third example is the blockchain, which is a process that protects people's wealth from being tampered with.

The trust in each system is usually based on a combination of those models. In benevolent dictatorships, for example, trust is mostly based on the good-heartedness of the leader. In democratic regimes, trust is more complex. It requires a combination of trusting the good-heartedness of whoever you want to elect, as well as trusting the process of separation of power, which prevents one person from having all the power to do whatever they wants. In addition, democratic regimes require a reelection process periodically, which provides the second type of trust; it is not in the interest of the elected person to do things that will jeopardize their chances of being reelected.

What is interesting about the blockchain, however, is that it does not use the first model of trust at all, and only uses the second model in edge cases.

In the blockchain, you do not have to trust anyone, you only trust the process. For all you know, the miners who are running the blockchain could be the most untrustworthy people out there. This is not a problem; the process is trustless. In the blockchain, the next block is mined by a miner who is chosen at random, based on who gets to solve the puzzle, which is -to a degree- a game of pure luck. And if you ever want to cheat the system, mining one block is not enough. You have to be the lucky miner several times in a row, which is almost impossible. So even if you intend to cheat, you won't be able to do that.

This is a huge move towards the mentality of "you know what, I am sick of trusting people, I will just assume everyone is bad, and design a process that will disallow them to harm me even if they try to". This stems from the fact that the first kind of trust could be easily manipulated by deception, media, fake news, etc. while the second kind can be broken by providing the trusted person with an interest that overcomes the disinterest that would result from breaking trust.

By fully trusting the process, you strip away the power from humans, and only trust the algorithm. There are some concerns to keep in mind here though:

•  Can you really design a system where you only have to trust the process? Bitcoin, for example, still has not reached that level. The part of bitcoin that keeps track of how much money everybody has, is indeed trustless, but there are other parts of the system that are not. For example, people have to trust the programmers responsible for bitcoin using the first or second model of trust. Most people cannot understand code or the technical details behind bitcoin, and so can only decide which bitcoin client to use based on their trust in the developing entity. If the developers of Bitcoin Core -the most used bitcoin client- decide to act maliciously, most people will keep following them without knowing.
• Processes are always hackable; the human race has yet to reach the level of designing loophole-free systems. There is always something that a hacker can use to take advantage of the system. Take, for example, the voting process in the Parliament, which is meant to prevent people from making decisions that only serve their interest. This process can be hacked by lobbying, by pulling strings, by blackmailing, and by making deals in order to make people vote for your interests. The process of bitcoin is also hackable; it is meant to protect people's wealth from tamperers. However, you can highly affect people's wealth by affecting the supply/demand of bitcoin; if you own a large sum of the currency, you could simply release it all at once into the market, which will decrease the price of bitcoin, and will also cause a frenzy among people, causing them to start selling, which will decrease the price even further.
• If you design a process that is 100% trustless, and it gets hacked, you will not be able to fix it without reverting back to other kinds of trust. Unless you design a self-fixing process, but I think we can prove mathematically that this is not doable, the same way we prove the undecidablity of the halting problem.
• When people trust a process for a long time, this process starts gaining a holy status in people's minds. Whenever the democratic process gets hacked by one of the ways we mentioned earlier, we see people saying: "Well, it is democracy, we tried to prevent that in a democratic way, but we failed. Hopefully it is for the best!". The democratic process is now deeply rooted in the society, that even its failures are considered a part of its holy existence, which decreases the general motivation to try and fix this process.

Thus, even though the blockchain is considered a huge step towards withdrawing confidence from people and placing it in the process, we need to realize that this is not yet entirely true, and that we might not really want to do that.