Overheard in my head

A: I'm thinking of the tip that authors of self help books go around giving people.
B: Which one?
A: When faced with a problem, ask yourself what someone else would do. Someone you admire and think is successful.
B: I always thought that was demeaning. It is as if you'd rather trust someone else's judgment than your own. Like the best you can do is to borrow another person's wisdom.
A: Yeah. I know. But that would be true if you were to actually walk up to the other person and ask for help. This is completely different. Don’t you see?
B: In the first case, you talk to the real person and in the second you seek the help of that person's model in your head. Sounds equally pathetic to me.
A: Think about it. You are consulting the other person's model in your head. Not the person. So, who is the one doing the thinking? You are running his software but its running on your machine. Makes sense?
B: Well, you are running his ready-made software. You have no idea how he made it. That's still no good.
A: Don’t push the analogy too far now. Forget the software example. You can run a program without knowing how it was written. But that's not true here. If you are consulting this person - I mean his phantom in your head - about a problem that you never saw the real person solve, and his phantom in your head produces a solution, isn't it you that really came up with it?
B: But then again, it is his algorithm that you are using to come up with the solution. It is not really yours. You blindly run your problem through his algorithm.
A: Now, that's a better analogy. And I agree with you. To run his algorithm in your head, you need to know his algorithm. And you wouldn’t be running it on your problem if you weren't convinced that it was right. So basically, you are using something that you understand well and something that you know to be a good general solution to the problem. Who cares if the algorithm was borrowed?
B: The point remains that it isn’t your solution. In fact, it's worse than that. You are probably convinced that his algorithm is right not because you can see why it works but because you think this person is successful and he cannot be so if his algorithms didn’t work. So basically, you are running an emulation of something you don’t understand because you have empirical evidence of its quality.
A: That's a hard one to refute. I'm probably oversimplifying this and that's why I'm losing. Let me start afresh. We know that we don't quite think algorithmically. And even if we do, having a clear understanding of someone else's algorithm - clear enough to emulate it - sounds impossible. Atleast extremely difficult. More often, we are guessing their algorithm. Based on our own understanding of everything. So unless we are actually emulating what the other person verbally claims to be doing, your knowledge of the other person's algorithm is nothing but your best explanation for how he does what ever he does. In fact it is your algorithm to produce his results. It is probably nothing like his algorithm. And probably is even way better than his.
B: If it's your algorithm, I wonder what the other person's contribution was at all. All he did was to assert that something was possible? Is that all we need?
A: It seems so. And that's not the end. We consult not one but multiple experts on different problems. Like a swiss army knife, we just pick whoever is best suited to the problem.
B: A swiss army knife of our models of peoples' wisdom. Strange but plausible. To go with it, probably there is a little bit of our original self that does the picking and choosing.
A: What is this swiss army knife but our own self? And isn't it as original as anything? Isn't each one of those phantom people nothing but our own doing? And do all of them have to be phantoms of real people? Can't you have a phantom of sherlock holmes? That would be a phantom of a phantom born in someone else's mind. An army knife of phantoms of real and phantom people. Sounds hardly like me.
B: A chimera of sorts. Interesting.