RE: Order by Protocol in Digital Space
I came here because of the scrabble picture. :D
The traditional model of order is not what is practiced in America. Yeah there is some order in that President > Vice Vp and Cabinet and there is a sort of chain in command. However, there are 2 additional branches of government that are equally powerful, but also have a chain of command. U.S. Supreme Court > Federal Appellate Courts > U.S. District Courts. (We taking federal only). We have actually 4 branches of government, if you think about it (regulatory Agencies).
I don't necessarily think knowledge gets lost at the top. How will a person or entity at the top do their job if they can't vet out the weaker data that could easily be screened by the lower peeps in the chain of command.
Also, with increasing algorithms creates new problems. You can code a program that can influence the behavior of a swarm, but what happens when they encounter a problem that is not addressed in the program? We still need to put a environment that can make those decisions. My point is that who will decide what limits, if any that the program has? Will the coder have ultimate say and how does that actually become a cyber-chain?
I literally learn something everyday. Thanks for information and making me think. As for the last post, I have no idea. :D
Current knowledge doesn't always reach the top. The governments whether US or otherwise all rely on intelligence agencies. These intelligence agencies give daily briefings or reports to the President so the President has decision support. So there is decision support for the President.
If you have to make a really big decision you can't expect to make it by yourself. You rely on your team of advisers. In the case of AI and blockchains these advisors would be digital.
The program is supposed to be able to evolve. When an unexpected problem is encountered then it evolves. Instead of using the word program instead try thinking about this as an autonomous agent. An autonomous agent like a bot, but this bot uses human beings for it's eyes, and it aggregates the human computation into a stream of numbers which get crunched and output comes out.
The point here is that the autonomous agent is just the AI portion of the system. It exists to do the computations that humans aren't good at. It does not replace human computation in the system. Human computation exists to do the computations that AI isn't good at. It's this symbiotic relationship which would enable it all to work and it's that symbiotic relationship that is missing from the "immutable code as law" ideal because without the ability to utilize human computation within your system you are restricting yourself to only being able to solve problems which can be solved by a program as you say.
The coders might come up with initial rules for a system but the coders do not direct the evolutionary trajectory. The evolutionary trajectory of these sorts of systems would be emergent and would be the result of continuous feedback from the human beings. Human preferences for example, human computations, big data, as a neuro network grows smarter if you have lots and lots of data to feed it.
So to put it simply, you do not hard code a governance system. You design the system for evolvability, to be able to adapt to change, to be resilient, to be flexible. As the participants change, it changes, as the users change, it changes, as social norms, culture, or preferences change, it changes.
I will address this more in future blog posts, sorry if there is lack of clarity or if it doesn't make sense right now.