2 ms·
The author does a good job of laying out the fundamental differences between the Jeffersonian and Hamitonian approaches to solving the "knowledge problem" of pl
by salawat 9y ago
The author does a good job of laying out the fundamental differences between the Jeffersonian and Hamitonian approaches to solving the "knowledge problem" of platforms creating such massive sets of data as to become a threat to any society in which they exist, but misses in terms of thinking out the consequences of several approaches.
For instance, there is a perverse disincentive in the Hamiltonian approach to relinquish control. The input which would "theoretically" increase the quality of a Hamiltonian regulator is more data. This means the regulator has a greater incentive to act as a data vacuum than those they are regulating. Furthermore, there comes a point at which when the data sets get large enough, there will tend to be no accountability to be found. See the banking crisis.
"Oops. We accidentally the economy. We promise we won't do it again."
"What did you even do?"
"We don't know off hand, but we won't do it again!"
The Jeffersonian approach is very much a precautionary approach. Where a Hamiltonian goes all in, the Jeffersonian steps back, limits scope, and sees what happens. It is better to advance the state of the art slowly and methodically without creating a point where a failure could propagate through the entire system, than to take the risk that the SCRAM (Safety Cut Rope Axe Man, the man responsible for triggering the active safety mechanisms in a non-passively safe nuclear reactor design) falls asleep at a bad time.
Do one thing, do it well. You don't need a Leviathan to build a functional, thriving society, but it sure helps if you are looking to destabilize and destroy one.