A reaction to omnizoid's latest post in this debate.
omnizoid proposed a problem in an earlier post, which I steelmanned as follows:
- there's an agent who can (almost) perfectly predict whether people will cut off their legs once they exist
- this agent only creates people who he predicts will cut off their legs once they exist
- existing with legs > existing without legs > not existing
My reaction to this problem was this:
FDT'ers would indeed cut off their legs: otherwise they wouldn't exist. omnizoid seems to believe that once you already exist, cutting off your legs is ridiculous. This is understandable, but ultimately false. The point is that your decision procedure doesn't make the decision just once. Your decision procedure also makes it in the predictor's head, when she is contemplating whether or not to create you. There, deciding not to cut off your legs will prevent the predictor from creating you.
omnizoid disagrees:
But you only make decisions after you exist. Of course, your decisions influence whether or not you exist but they don’t happen until after you exist. But once you exist, no matter how you act, there is a zero percent chance that you won’t exist.
But if the predictor implemented your decision procedure when making his prediction - and this is, I believe, what is usually meant by "predicting" in this context - then your decision is NOT only made after you exist: it is also made in the predictor's head before you even exist. That is the whole point of FDT: subjunctive dependence.
omnizoid and I also debated Wolfgang Schwarz's Procreation example. In omnizoid's latest post, he does mention the problem, but doesn't engage with my latest point - instead, he basically just reiterates his latest point. Therefore, I won't further discuss this part.
Unfortunately, omnizoid does continue his arguing-from-authority thing:
I think higher-order evidence of this kind is useful. If 99.9% of people who seriously study a topic agree about X, and you disagree about X, you should think your reasoning has gone wrong.
I actually agree this is a pretty good indication you are wrong about X. However, to make this particular claim, you should
omnizoid has done neither, so his claim is pretty much meaningless. (If he had done both, I would still not reject FDT, as I believe I have good reasons for believing it is a good theory.)
Moving on to MacAskill's implausible discontinuities objection:
A related problem is as follows: FDT treats ‘mere statistical regularities’ very differently from predictions. But there’s no sharp line between the two. So it will result in implausible discontinuities. There are two ways we can see this.
First, take some physical processes S (like the lesion from the Smoking Lesion) that causes a ‘mere statistical regularity’ (it’s not a Predictor). And suppose that the existence of S tends to cause both (i) one-boxing tendencies and (ii) whether there’s money in the opaque box or not when decision-makers face Newcomb problems. If it’s S alone that results in the Newcomb set-up, then FDT will recommending two-boxing.
But now suppose that the pathway by which S causes there to be money in the opaque box or not is that another agent looks at S and, if the agent sees that S will cause decision-maker X to be a one-boxer, then the agent puts money in X’s opaque box. Now, because there’s an agent making predictions, the FDT adherent will presumably want to say that the right action is one-boxing. But this seems arbitrary — why should the fact that S’s causal influence on whether there’s money in the opaque box or not go via another agent much such a big difference?
I objected to this critique by saying that MacAskill is wrong here:
the critical factor is not whether "there's an agent making predictions". The critical factor is subjunctive dependence, and there is no subjunctive dependence between S and the decision maker here.
But omnizoid insisted, and insist again, that there is subjunctive dependence. This time, he writes:
The agent looks at the brain and then acts only if the brain is likely to output the result of one boxing.
No, the agent looks at the lesion, not at the brain. omnizoid and I have chatted about this a bit, and we could assume that the agent who makes the prediction can infer your decision procedure in this problem just by observing whether S is or isn't present. In that case, the agent could implement your decision procedure, and then there would be subjunctive dependence. I think this is a pretty absurd assumption, but whatever. If we assume this, there is subjunctive dependence. But then, MacAskill's entire objection doesn't hold: the problem has become equivalent to Newcomb's Problem. There is nothing implausible here: without the agent making the prediction, the problem was like Smoking Lesion. With the agent, the problem is like Newcomb's Problem. The supposed implausible discontinuity has become the absence or presence of subjunctive dependence itself. Which, of course, you can object to, but calling it an implausible discontinuity wouldn't do much. (Btw, I don't think this is what MacAskill had in mind either.)
To be clear, I maintain there is no subjunctive dependence. But with or without it, MacAskill has no argument against FDT.
That's it for this post. Thanks for reading!