4 ms·
There are plenty of examples where one party or a collation of parties set up the rules for a decentralized system that then goes on to operate successfully wit
by jcoffland 9y ago
There are plenty of examples where one party or a collation of parties set up the rules for a decentralized system that then goes on to operate successfully with little to no further enforcement. Email or TCP/IP are examples of such decentralized systems.
The fact that someone must make the rules and future actors may attempt to change those rules are problems but they do not invalidate the idea of decentralized systems.
I believe the author is reacting to his colleges misunderstanding of crypto technologies. He has bought in to their argument that what cryptocurrency needs to solve it's current problems is centralized control or governance. This, of course leads to the logical conclusion that if you must have centralized control then why use a blockchain at all.
The fallacy is in accepting that after the initial rules are set, that you must continue to have governance to solve the inevitable disagreements. On the contrary, continued governance is the problem faced by cryptocurrencies and other blockchain technologies. Their stability depends on adherence to the original agreement. Any change to that agreement is an existential threat.
It would however be a mistake to think that this problem completely invalidates such decentralized systems. They can and do continue to provide utility even in the face of attempts at continued governance. It's not about these systems working for eternity. It's about them working for long enough for life to happen.
- jasode 9y ago>Email or TCP/IP are examples of such decentralized systems. I'm not sure about those analogies. SMTP email and TCPIP require DNS to route packets. DNS root authority is centralized. (Yes, DNS for performance purposes is implemented across distributed computers worldwide but ultimate authority still traces back to the root servers.) Even without DNS lookup, the numerical IP addresses are assigned by a central authority. (Yes, if we're only talking about playing with protocols instead of actual global connectivity, one could run private SMTP on an isolated network with no central authority but that's not the analogy we're interested in.) Maybe your analogies and my disagreement with them is a perfect illustration of what the author is talking about: centralized governance is still there even if we don't notice it.
- runeks 9y agoIt's not about eliminating centralization, it's about minimizing the fragility it causes in key areas. We minimize it by locating bottlenecks and finding solutions in these areas, not throwing our hands in the air unless we have 100% decentralization (assuming we even want that). Sensible proponents of Bitcoin shouldn't argue that centralization is "wrong", but that monetary centralization creates perverse incentives, and that Bitcoin is a potential solution to this.