# pragmatist

Karma: 4,969
• At least some of the arguments offered by Richard Rorty in Philosophy and the Mirror of Nature are great. Understanding the arguments takes time because they are specific criticisms of a long tradition of philosophy. A neophyte might respond to his arguments by saying “Well, the position he’s attacking sounds ridiculous anyway, so I don’t see why I should care about his criticisms.” To really appreciate and understand the argument, the reader needs to have sense of why prior philosophers were driven to these seemingly ridiculous positions in the first place, and how their commitment to those positions stems from commitment to other very common-sensical positions (like the correspondence theory of truth). Only then can you appreciate how Rorty’s arguments are really an attack on those common-sensical positions rather than some outre philosophical ideas.

• Perhaps explicitly thinking of them as systems of equations (or transformations on a vector) would be helpful.

As an example, suppose you are asked to multiply matrices A and B, where A is [1 2, 0 4, −1 2] (the commas represent the end of a row) and B is [2 1 0, 3 1 2]. Start out by taking the rightmost matrix (B in this case) and converting it into a series of equations, one for each row. So since the first row is 2 1 0, the relevant equation will be 2x + 1y + 0z. Assign each of these equations to some other variable. So we now have

X = 2x + y

Y = 3x + y + 2z

Now do the same thing with the matrix on the left, except this time use the new variables you’ve introduced (X and Y), so the three equations you end up with (one for each row) will be

X + 2Y

4Y

-X + 2Y

Now that you have these formulae, substitute in the values of X and Y based on your earlier equations. You get

(2x + y) + 2(3x + y + 2z)

4(3x + y + 2z)

-(2x + y) + 2(3x + y + 2z)

Simplifying, you get

8x + 3y + 4z

12x + 4y + 8z

4x + y + 4z

The coefficients of these equations are the result of the multiplication. So the product of the two matrices is [8 3 4, 12 4 8, 4 1 4].

I’ll admit this is not the quickest way to go about multiplying matrices, but it might be easier for you to remember since it doesn’t seem as arbitrary. And maybe once you get used to thinking about multiplication this way, the usual visual rule will start making more sense to you.

• I think Bostrom’s argument applies even if they aren’t “highly accurate”. If they are simulated at all, you can apply his argument.

I don’t think that’s true. The SSA will have different consequences if the simulated minds are expected to be very different from ours.

If we suppose that simulated minds will have very different observations, experiences and memories from our own, and we consider the hypothesis that the vast majority of minds in our universe will be simulated, then SSA simply disconfirms the hypothesis. If I should reason as if I am a random sample from the pool of all observers, then any theory which renders my observations highly atypical will be heavily disconfirmed. SSA will simply tell us it is unlikely that the vast majority of minds are simulated. Which means that either civilizations don’t get to the point of simulating minds or they choose not to run a significant number of simulations.

If, on the other hand, we suppose that a significant proportion of simulated minds will be quite similar to our own, with similar thoughts, memories and experiences, and we further assume that the vast majority of minds in the universe are simulated, then SSA tells us that we are likely simulated minds. It is only under those conditions that SSA delivers this verdict.

This is why, when Bostrom describes the Simulation Argument, he focuses on “ancestor-simulations”. In other words, he focuses on post-human civilizations running detailed simulations of their evolutionary history, not just simulations of any arbitrary mind. It is only under the assumption that post-human civilzations run ancestor-simulations that the SSA can be used to conclude that we are probably simulations (assuming that the other two possible conclusions of the argument are rejected).

So i think it matters very much to the argument that the simulated minds are a lot like the actual minds of the simulators’ ancestors. If not, the argument does not go through. This is why I said you seem to simply be accepting (2), the conclusion that post-human civilizations will not run a significant number of ancestor-simulations. Your position seems to be that the simulations will probably be radically dissimilar to the simulators (or their ancestors). That is equivalent to accepting (2), and does not conflict with the simulation argument.

You seem to consider the Simulation Argument similar to the Boltzmann brain paradox, which would raise the same worries about empirical incoherence that arise in that paradox, worries you summarize in the parent post. The reliability of the evidence that seems to point to me being a Boltzmann brain ts itself predicated on me not being a Boltzmann brain. But the restriction to ancestor-simulations makes the Simulation Argument importantly different from the Boltzmann brain paradox.

• I am taking issue with the conclusion that we are living in a simulation even given premise (1) and (2) being true.

(1) and (2) are not premises. The conclusion of his argument is that either (1), (2) or (3) is very likely true. The argument is not supposed to show that we are living in a simulation.

He’s saying that (3) doesn’t hold if we are not in a simulation, so either (1) or (2) is true. He’s not saying that if we’re not in a simulation, we somehow are actually in a simulation given this logic.

Right. When I say “his conclusion is still true”, I mean the conclusion that at least one of (1), (2) or (3) is true. That is the conclusion of the simulation argument, not “we are living in a simulation”.

If I conclude that there are more simulated minds than real minds in the universe, I simply do not think that implies that I am probably a simulated mind.

This, I think, is a possible difference between your position and Bostrom’s. You might be denying the Self-Sampling Assumption, which he accepts, or you might be arguing that simulated and unsimulated minds should not be considered part of the same reference class for the purposes of the SSA, no matter how similar they may be (this is similar to a point I made a while ago about Boltzmann brains in this rather unpopular post).

I actually suspect that you are doing neither of these things, though. You seem to be simply denying that the minds our post-human descendants will simulate (if any) will be similar to our own minds. This is what your game AI comparisons suggest. In that case, your argument is not incompatible with Bostrom’s conclusion. Remember, the conclusion of the simulation argument is that either (1), (2), or (3) is true. You seem to be saying that (2) is true—that it is very unlikely that our post-human descendants will create a significant number of highly accurate simulations of their descendants. If that’s all you’re claiming, then you’re not disagreeing with the simulation argument.

• First, Bostrom is very explicit that the conclusion of his argument is not “We are probably living in a simulation”. The conclusion of his argument is that at least one of the following three claims is very likely to be true -- (1) humans won’t reach the post-human stage of technological development, (2) post-human civilizations will not run a significant number of simulations of their ancestral history, or (3) we are living in a simulation.

Second, Bostrom has addressed the objection you raise here (in his Simulation Argument FAQ, among other places). He essentially flips your disjunctive reasoning around. He argues that we are either in a simulation or we are not. if we are in simulation, then claim 3 is obviously true, by hypothesis. If we are not in a simulation, then our ordinary empirical evidence is a veridical guide to the universe (our universe, not some other universe). This means the evidence and assumptions used as the basis for the simulation argument are sound in our universe. It follows that since claim 3 is false by hypothesis, either claim 1 or claim 2 is very likely to be true. It’s worth noting that these two are claims about our universe, not about some parent universe.

In other words, your objection is based on the argument that if we are in a simulation, there is no good reason to trust the assumptions of the simulation argument (such as assumptions about how our simulators will behave). Bostrom’s reply is that if we are in a simulation, then his conclusion is true anyway, even if the specific reasoning he uses doesn’t apply. If we are not in a simulation, then the reasoning he uses does apply, so his conclusion is still true.

There does seem to be some sort of sleight-of-mind going on here, if you want my opinion. I generally feel that way about most non-trivial uses of anthropic reasoning. But the exact source of the sleight is not easy for me to detect. At the very least, Bostrom has a prima facie response to your objection, so you need to say something about why his response is flawed. Making your objection and Bostrom’s response mathematically precise would be a good way to track down the flaw (if any).

• “Why is it that when our cognitive systems engage in visual and auditory information-processing, we have visual or auditory experience: the quality of deep blue, the sensation of middle C? How can we explain why there is something it is like to entertain a mental image, or to experience an emotion? It is widely agreed that experience arises from a physical basis, but we have no good explanation of why and how it so arises. Why should physical processing give rise to a rich inner life at all?”

-- David Chalmers

These questions may be a product of conceptual confusion, but they don’t seem that way to me. Perhaps I am confused in the same way.

• When you update, you’re not simply imagining what you would believe in a world where E was true, you’re changing your actual beliefs about this world. The point of updates is to change your behavior in response to evidence. I’m not going to change my behavior in this world simply because I’m imagining what I would believe in a hypothetical world where E is definitely true. I’m going to change my behavior because observation has led me to change the credence I attach to E being true in this world.

• Updating by Bayesian conditionalization does assume that you are treating E as if its probability is now 1. If you want an update rule that is consistent with maintaining uncertainty about E, one proposal is Jeffrey conditionalization. If P1 is your initial (pre-evidential) distribution, and P2 is the updated distribution, then Jeffrey conditionalization says:

P2(H) = P1(H | E) P2(E) + P1(H | ~E) P2(~E).

Obviously, this reduces to Bayesian conditionalization when P2(E) = 1.

• Credit and accountability seem like good things to me, and so I want to live in a world where people/​groups receive credit for good qualities, and are held accountable for bad qualities.

If this is your concern, then you should take into account what sorts of groups are appropriate loci for credit and accountability. This will, of course, depend on what you think is the point of credit/​accountability.

If you believe, as I do, that the function of credit and accountability is to influence future behavior, then it seems that the appropriate loci of credit/​accountability should be “agential”. In other words, objects of credit and blame should be capable of something resembling goal-directed alteration of behavior. Individual people are appropriate loci on this account, since they are (at least, mostly) paradigmatic agents.

Some groups might also qualify as agential, and thus as appropriate loci of credit and blame. Corporations come to mind, as do nations. But that is because those groups have a particular organizational structure that makes them somewhat agent-like. Not every group has this quality. The group of all left-handed people, for instance, is not agent-like in any relevant sense, so I don’t see the point of assigning credit or blame to it. Similarly for racial groups or genders.

• It seems to me that your objection here is driven mainly by a general dislike of Gleb’s contributions (and perhaps his presence on LW), rather than a sincere conviction about the importance of your point. I mean, this is a ridiculous nitpick, and the hostility of your call-out is completely disproportionate to the severity of Gleb’s supposed infraction.

While Gleb’s aside might be a “lie” by some technical definition, it certainly doesn’t match the usual connotations of the term. I see virtually zero harm in the kind of “lie” you’re focusing on here, so I’m not sure about the value of your piece of advice, other than signalling your aversion towards Gleb.

• I absolutely agree that Kant’s system as represented in the Groundwork is unworkable. But you could say the same about pretty much any pre-20th-century philosopher’s major work. I think the fact that someone was even trying to think about ethics along essentially game-theoretic lines in the 18th century is pretty revolutionary and worthy of respect, even if he did get important things wrong. As far as I’m aware, no one else was even in the ballpark.

ETA: I do think a lot of philosophers scoff (correctly) at Kant’s object-level moral views, not only because of their absurdity (the horrified tone in which he describes masturbation still makes me chuckle) but because of the intellectual contortions he would go through to “prove” them using his system. While I believe he made very important contributions to meta-ethics, his framework was nowhere near precise enough to generate a workable applied ethics. So yeah, Kant’s actual ethical positions are pretty scoff-worthy, but the insight driving his moral framework is not.

• Isn’t that motte/​bailey

Not sure it’s motte-and-bailey. I do think there are several serious pathologies in large swathes of contemporary philosophy. And I say this not as a dilettante, but a professional philosopher. There are areas of philosophy where these pathological tendencies are being successfully held at bay, and I do think there are promising signs that those areas are growing in influence. But much of mainstream philosophy, especially mainstream metaphysics and epistemology, does suffer from continued adherence to what I consider archaic and unhelpful methodology. And I think that’s what Luke is trying to point out. He does go overboard with his rhetoric, and I think he lacks a feel for the genuine insights of the Western philosophical tradition (as smart and insightful as I think Yudkowsky is, I really find it odd that someone who purports to be reasonably familar with philosophy would cite him as their favorite philosopher). But I think there is a sound point lurking under there, and not merely a banal “motte”-style point.

I was only pointing out in response to the OP that I have been harping on LW’s silly anti-academic sentiment for ages, that’s all.

I absolutely agree with you on the silliness of the anti-academic sentiment.

• I think Luke will agree with you on what you say here, though. I remember commenting on one of his posts that was critical of philosophy, saying that his arguments didn’t really apply to the area of philosophy I’m involved in (technical philosophy of science). Luke’s response was essentially, “I agree. I’m not talking about philosophy of science.” I think he’d probably say the same about philosophical work on decision theory and causal inference.

• I don’t know of many analytic philosophers who scoff at his ethics, although there are certainly many who disagree with it. There are also many analytic philosophers who consider his ethics to be a significant advance in moral reasoning. As an example, Derek Parfit, in his recent book, constructs an ethical system that tries to reconcile the attractions of both consequentialism and Kantian deontological ethics.

Kant’s discussion of the categorical imperative, especially the first formulation of the imperative (act according to the maxim that you would will to be a universal law), prefigures various contemporary attempts to reformulate decision theory in order to avoid mutual defection in PD-like games, including Hofstadter’s notion of superrationality and Yudkowsky’s Timeless Decision Theory. Essentially, Kantian ethics is based on the idea that ethics stems from nothing more than a commitment to rational decision-making and common knowledge of rationality among interacting agents (although with Kant it’s not so much about knowing that other agents are rational but about respecting them by treating them as rational). I don’t fully agree with this perspective, but I do think it is remarkably astute and ahead of its time.

• Kant is talking about good and evil, delight, happiness, character, honor, etc., etc, while Russell is talking about looking at triangles. Which one are people going to want to read?

Except Kant also talked quite a bit about triangles and Russell also talked quite a bit about good and evil. And Kant discussed perceptual epistemology a whole lot more than Russell did. The Critique of Pure Reason, Kant’s most significant work, is about epistemology, not ethics.

Also, while much of twentieth-century continental philosophy does build on Kant (although a lot of it is a reaction against Kant), so does much of twentieth-century analytic philosophy. In many ways, the true heirs of Kant in the twentieth century were the logical positivists. Their epistemology was closer to Kant’s than any prominent continental philosopher’s was. So Kant has just as much claim to being a founder of analytic philosophy as he does to being a founder of continental philosophy.

Kant was not the most lucid writer, but his style was not remotely “analogical” or “literary” (look through Kant’s famous Transcendental Deduction and see whether those descriptors seem apt).. And much of Kantian philosophy is precisely formulated and subject to falsification. In fact, quite a bit of it has been falsified (his contention that space is necessarily Euclidean, for instance).

• Relativists have no non-subjective notion of “normativity”, thus the subjective/​normative distinction makes no sense to them.

This is not true of all relativists. There are relativists who believe in entirely objective agent-relative moral facts. In other words, they would say something like, “It is an objective moral truth that X is wrong for members of community Y”. The normative force of “X is wrong” would apply even to members of community Y who don’t believe that X is wrong (hence the objectivity), but it wouldn’t apply to people outside community Y (hence the relativism).

• Mordecai Kaplan would be unhappy to hear that commitment to ritual and tradition requires belief

I think the issue is not whether commitment to ritual—as in, a commitment to go through the motions—requires belief, it’s whether experiencing ritual as beautiful requires belief. I think it’s plausible that immersing oneself in the context of the ritual, including the requisite belief set, makes it far more meaningful and awe-inspiring. Merely aesthetic appreciation of ritual may not inspire the same depth of feeling as you would experience if every move in the ritual were wrought with spiritual significance for you.

So participating in the tradition without believing may also count as “depriving oneself of beauty”. I wouldn’t really know, though. I’ve been a non-believer my entire intellectually aware life, so I have no basis for comparison. I will say that I can’t imagine any ritual or tradition driving me into the kind of frenzy you see at some charismatic Pentecostal churches, for instance. But I can’t really imagine being driven to the kind of frenzy you see in the average audience for the The Price is Right either, so this may be an issue of personality rather than belief.

• Boltzmann’s original combinatorial argument already presumed a discretization of phase space, derived from a discretization of single-molecule phase space, so we don’t need to incorporate quantum considerations to “fix” it. The combinatorics relies on dividing single-particle state space into tiny discrete boxes, then looking at the number of different ways in which particles could be distributed among those boxes, and observing that there are more ways for the particles to be spread out evenly among the boxes than for them to be clustered. Without discretization the entire argument collapses, since no more than one particle would be able to occupy any particular “box”, so clustering would be impossible.

So Boltzmann did successfully discretize a box full of particles with arbitrary position and momentum, and using his discretization he derived (discrete approximations of) the Maxwell-Boltzmann distribution and the Boltzmann formula for entropy. And he did all this without invoking (or, indeed, being aware of) quantum considerations. So the Sackur-Tetrode route is not a requirement for a discretized Boltzmann-esque argument. I guess you could argue that in the absence of quantum considerations there is no way to justify the discretization, but I don’t see why not. The discretization need not be interpreted as ontological, emerging from the Uncertainty Principle. It could be interpreted as merely epistemological, a reflection of limits to our abilities of observation and intervention.

Incidentally, none of these derivations require the assumption of ergodicity in the system. The result that the size of a macrostate in phase space is proportional to the number of microstates emerges purely from the combinatorics, with no assumptions about the system’s dynamics (other than that they are Hamiltonian). Ergodicity, or something like it, is only required to establish that the time spent by a system in a particular macrostate is proportional to the size of the macrostate, and that is used to justify probabilistic claims about the system, such as the claim that a closed system observed at an arbitrary time is overwhelmingly likely to be in the macrostate of maximum Boltzmann entropy.

So ultimately, I do think the point Carroll was making is valid. The Boltzmann entropy—as in, the actual original quantity defined by Boltzmann and refined by the Ehrenfests, not the modified interpretation proposed by people like Jaynes—is distinct from the Gibbs entropy. The former can increase (or decrease) in closed system, the latter cannot.

To put it slightly more technically, the Gibbs entropy, being a property of a distribution that evolves according to Hamiltonian laws, is bound to stay constant by Liouville’s theorem, unless there is a geometrical change in the accessible phase space or we apply some coarse-graining procedure. Boltzmann entropy, being a property of macrostates, not of distributions, is not bound by Liouville’s theorem. Even if you interpret the Boltzmann entropy as a property of a distribution, it is not a distribution that evolves in a Hamiltonian manner. It evolves discontinuously when the system moves from one Boltzmann macrostate to the next.