Some thoughts on the Belief-Machine

So, my noble interlocutor Jack finds the belief-machine "sinister." I agree that there would be something fishy about using the machine. The point of the idea is to pinpoint the fishiness. So, let's go.

One might say that using the machine subverts one's autonomy (or free will). But, paradoxically perhaps, sometimes "giving up control," or "letting go," is precisely what's needed to liberate ourselves from unnecessary suffering. Of course, one might say that's exactly the problem with the machine: using it is a hyperactive effort to control our own beliefs. But if we were to imagine someone who really wanted to believe something (say, an extreme pessimist who wanted to believe in the goodness of people, or just that goodness is possible...), we might think that the machine would be a quick way to relieve this problem.

A worry I have is the thought of having a belief of significance with no sense of why one believes it. One would, of course, be free to seek out reasons, but we might worry that such an inquiry would be biased from the get-go, pure confabulation. It might be thought that one's inability to adopt a particular belief on one's own is reason for thinking that one shouldn't have the belief. But the sort of beliefs I'm imagining one might have "put in" are ones that others have, and I might not understand their reasons, but trust their judgment, and so think I should believe that way, too, even though I can't bring myself to do it.

Certainly there are things about which I should trust others more than myself, and I might think of my failure to do so as a failure of character. Now, you might say that the problem with the belief-machine in that case is that it addresses a symptom rather than the "disease": it "corrects" my belief, but not the character flaw, as I see it, that prevents me from believing as I think I should.

So, is that the problem: not that the machine circumvents the need to understand why one believes such-and-such, but rather that using the machine involves circumventing the problem of understanding oneself?

The Belief-Machine

Suppose there were a belief-machine that could make you believe whatever you wanted to believe. There are sometimes cases where, intellectually, we find a particular claim compelling, but our "heart," as it were, isn’t in it, and thus we can’t bring ourselves to believe it. Thus, we have what William James called "a divided self." If we want to have religion, the belief-machine can give us religious conviction. If we want to believe that eating meat is wrong, we’ll really believe it. Weakness of will notwithstanding, the belief-machine would have the further effect of changing how we behave. So upon acquiring the belief that eating meat is wrong, we would have motivation to stop eating meat (like we never had before). One thing the belief-machine does not do is provide us with reasons for having these beliefs; they simply seem "natural" or "intuitive" to us. (That doesn't presumably prevent us from coming up with reasons for our beliefs later on.) My question is: would using the belief-machine be wrong? If so, why? (We say it is wrong for others to manipulate unwilling persons, but if I am willing to be manipulated, then what’s wrong with using the belief-machine?)

(Obviously, kudos to Robert Nozick for machine-type thought experiments of this sort.)