I imagine something like teaching people a “think about the problem—gather data—make a decision—do it” loop, starting with simple problems and gradually moving more meta.
The emphasis would be on doing; on closing all the started loops. At the beginning there would be dozens of simple problems, all of them have to be solved very soon. Not just hear the theory and profess an agreement, but repeat it so many times it becomes your natural reaction. This would create a habit and an expectation that problems can be solved without procrastination. (Something like the “success spiral” in motivation literature.) Then move to more complex things, and perhaps add more steps to the loop, like “reflect upon what happened”. The idea would be to train people out of feeling helpless and stopping (i.e. not merely stopping doing, but also not gathering data, not discussing the problem, etc.).
I imagine that when a person with this training would see something that can be fixed easily, the person would a) recognize it as belonging to the class of the problems that can be fixed easily, b) make a cost-benefit decision, mostly about the opportunity costs, and in case of a positive decision, c) fix the problem. And the next problem, and the next problem. If you had this person as a friend, your observation from outside would be that many problems just somehow disappear when this person walks around. Then, if you were interested, there is a chance this person would think about a way to teach you this skill efficiently, too.
I imagine that given our human limitations, an important part of the skill would be to have a supervisor or a support group, and regularly report to them your problems and solutions; sharing knowledge and encouraging each other.
I would expect that something similar already exists, but it’s very context-specific, not generalized. For example how to behave as a spy or a soldier, or how to survive in a jungle… with a lot of domain-specific knowledge, and some training to solve problems soon and not procrastinate. But there is a dichotomy between being in the specific context, and being “outside the lab”. Perhaps this dichotomy is even considered important for the sanity of the participant; we don’t want people to behave in everyday circumstances the same way they would in a jungle or in the enemy territory. But for an x-rationalist trying to optimize the universe, the whole unoptimized world is a jungle, and the world where people keep dying in thousands every second is a hostile territory.
What would you do specifically if you wanted to teach agentiness?
I imagine something like teaching people a “think about the problem—gather data—make a decision—do it” loop, starting with simple problems and gradually moving more meta.
The emphasis would be on doing; on closing all the started loops. At the beginning there would be dozens of simple problems, all of them have to be solved very soon. Not just hear the theory and profess an agreement, but repeat it so many times it becomes your natural reaction. This would create a habit and an expectation that problems can be solved without procrastination. (Something like the “success spiral” in motivation literature.) Then move to more complex things, and perhaps add more steps to the loop, like “reflect upon what happened”. The idea would be to train people out of feeling helpless and stopping (i.e. not merely stopping doing, but also not gathering data, not discussing the problem, etc.).
I imagine that when a person with this training would see something that can be fixed easily, the person would a) recognize it as belonging to the class of the problems that can be fixed easily, b) make a cost-benefit decision, mostly about the opportunity costs, and in case of a positive decision, c) fix the problem. And the next problem, and the next problem. If you had this person as a friend, your observation from outside would be that many problems just somehow disappear when this person walks around. Then, if you were interested, there is a chance this person would think about a way to teach you this skill efficiently, too.
I imagine that given our human limitations, an important part of the skill would be to have a supervisor or a support group, and regularly report to them your problems and solutions; sharing knowledge and encouraging each other.
I would expect that something similar already exists, but it’s very context-specific, not generalized. For example how to behave as a spy or a soldier, or how to survive in a jungle… with a lot of domain-specific knowledge, and some training to solve problems soon and not procrastinate. But there is a dichotomy between being in the specific context, and being “outside the lab”. Perhaps this dichotomy is even considered important for the sanity of the participant; we don’t want people to behave in everyday circumstances the same way they would in a jungle or in the enemy territory. But for an x-rationalist trying to optimize the universe, the whole unoptimized world is a jungle, and the world where people keep dying in thousands every second is a hostile territory.