> Without a single trusted party. You're trading trust for one entity for trust for many entities.
The notion of “single” entities in society is, I think, flawed. A bank is not a single entity, for instance—they are not just a single actor—they are governed by a board and bylaws and shareholders and government regulations and judges, etc. If something goes wrong, which is expected and built into the system process,there is due process. And the strength of that due process is governed by the self interest of all the parties at play.
The main purpose of blockchain is to attempt to replace the messiness of human governance with the certainty of algorithmic governance. But it is an illusion in many ways and dangerous, in a sense, to put so much trust in an algorithm. I think that much of the philosophy around blockchain comes from the lack of faith in human governance—which is to say, a certain kind of misanthropy.
> The main purpose of blockchain is to attempt to replace the messiness of human governance with the certainty of algorithmic governance.
Right, and as is shown inevitably in real life (The DAO, BTC 0.7/0.8 fork, censoring of scam or Russia related BTC addresses, ...), the messiness of reasonably well understood and regulated human governance (law, elections, ...) is replaced by governance of the algorithm plus a bunch of unelected, unaccountable, and largely even unknown humans.
> I think that much of the philosophy around blockchain comes from the lack of faith in human governance—which is to say, a certain kind of misanthropy.
So the blockchain gets two things right: avoiding double-spending and an honest assessment on human nature.
Replacing trust in humanity with trust in an algorithm created by humanity doesn't really get you anywhere. Like it or not, you are human, I am human, and we live a society of humans! Everything in society is created and run by humans! The blockchain cannot change this fundamental fact, it can only obscure it.
You have to trust the people who figured out the math, when they say "this math is secure". Then you have to trust the people who programmed that math, when they say "this code implements this math correctly". And then you have to trust the binary, unless you compiled it yourself, when it says "this binary compiles this code correctly", and finally you have to trust the machines when they say "this computer executes this binary correctly", all the while trusting youself not to have made any mistakes at any step which will quickly cost you everything you put into the system.
If you are an expert mathematician, programmer, electrical engineer, and white hat you might not have to trust anyone else. How many people are all of these though?
> You have to trust the people who figured out the math, when they say "this math is secure".
Not really. I'm convinced the math is secure given the current knowledge, because there's a massive financial incentive to break it. I don't have to trust anyone.
