= 27d567bc2d48d9f40e9ccc20ea370b15)
Will_Sawin comments on Less Wrong Rationality and Mainstream Philosophy - Less Wrong
You are viewing a comment permalink. View the original post to see all comments and the full post content.
= 783df68a0f980790206b9ea87794c5b6)
Loading…= b715d02e85f7b3b4ae65ca3d3e450868)
Subscribe to RSS Feed
Powered by Reddit
= f037147d6e6c911a85753b9abdedda8d)
You are viewing a comment permalink. View the original post to see all comments and the full post content.
Comments (328)
From my small but nontrivial knowledge of Quine, he always struck me as having a critically wrong epistemology.
LW-style epistemology looks like this:
whereas Quine's seems more like
which seems to be missing most of the point.
His boat model always struck me as something confused that should be strongly modified or replaced by a Bayesian epistemology in which posterior follows logically and non-destructively from prior, but I may be in the minority in LW on this.
It's true that Quine lacked the insights of contemporary probability theory and AI, but remember that Quine's most significant work was done before 1970. Quine was also a behaviorist. He was wrong about many things.
My point was that both Quine and Yudkowsky think that recursive justification bottoms out in using the lens that sees its own flaws to figure out how humans gain knowledge, and correcting mistakes that come in. That's naturalized epistemology right there. Epistemology as cognitive science. Of course, naturalized epistemology has made a lot of progress since then thanks to the work of Kahneman and Tversky and Pearl and so on - the people that Yudkowsky learned from.
Bayesian inference is not a big step up from Laplace, and the idea of an optimal model that humans should try to approximate is a common philosophical position.
Who cares when his work was done. We want to know how to find work that helps us to understand things today. It's not about how smart he was, but about how much his ideas can help us.
And my answer is "not much." Like I say, all the basics of Quinean philosophy are already assumed by Less Wrong. I don't recommend anyone read Quine. It's (some of) the stuff his followers have done in the last 30 years that is useful - both stuff that is already being used by SIAI people, and stuff that is useful but (previously) undiscovered by SIAI people. I listed some of that stuff here.
What's wrong with behaviorism? I was under the impression that behaviorism was outdated but when my daughter was diagnosed as speech-delayed and borderline autistic we started researching therapy options. The people with the best results and the best studies (those doing 'applied behavior analysis') seem to be pretty much unreconstructed Skinnerists. And my daughter is making good progress now.
I'll take flawed philosophy with good results over the opposite any day of the week. But I'm still curious about flaws in the philosophy.
May I recommend Dennett's "Skinner Skinned", in Brainstorms?
Okay, I read it. It's funny how Dennett's criticism of Skinner partially mirrors Luke's criticism of Eliezer. Because Skinner uses terminology that's not standard in philosophy, Dennett feels he needs to be "spruced up".
"Thus, spruced up, Skinner's position becomes the following: don't use intentional idioms in psychology" (p. 60). It turns out that this is Quine's position and Dennett sort of suggests that Skinner should just shut up and read Quine already.
Ultimately, I can understand and at least partially agree with Dennett that Skinner goes too far in denying the value of mental vocabulary. But, happily, this doesn't significantly alter my belief in the value of Skinner type therapy. People naturally tend to err in the other direction and ascribe a more complex mental life to my daughter than is useful in optimizing her therapy. And I still think Skinner is right that objections to behaviorist training of my daughter in the name of 'freedom' or 'dignity' are misplaced.
Anyway, this was a useful thing to read - thank you, ciphergoth!
Thank you, holding the book in my hand and reading it now.
No, I'm talking about behaviorist psychology. Behaviorist psychology denied the significance (and sometimes the existence) of cognitive states. Showing that cognitive states exist and matter was what paved the way to cognitive science. Many insights from behaviorist psychology (operant conditioning) remain useful, but it's central assumption is false, and it must be false for anyone to be doing cognitive science.
Okay, but now I'm getting a bit confused. You seem to me to have come out with all the following positions:
Those things don't seem to go well together. What am I misunderstanding?
Quinean naturalism does not have an exclusive lock on useful philosophy, but it's the most productive because it starts from a bunch of the right assumptions (reductionism, naturalized epistemology, etc.)
Like I said, Quine was wrong about lots of things. Behaviorism was one of them. But Quine still saw epistemology as a chapter of the natural sciences on how human brains came to knowledge - the field we now know as "cognitive science."
Quine apparently said, "I consider myself as behavioristic as anyone in his right mind could be". That sounds good, can I subscribe to that?
Personally, I'm finding that avoiding anthropomorphising humans, i.e. ignoring the noises coming out of their mouths in favour of watching their actions, pays off quite well, particularly when applied to myself ;-) I call this the "lump of lard with buttons to push" theory of human motivation. Certainly if my mind had much effect on my behaviour, I'd expect to see more evidence than I do ...
I take exception to that: I have a skeletal structure, dammit!
I think the reference is to the brain rather than to the whole body.
(blink)
(nods) Yes, indeed.
Exception withdrawn.
Well played!
It sounds like what you are describing is rationalization, either doing it yourself or accepting people's rationalization about themselves.
Pretty much. I'm saying "mind" for effect, and because people think the bit that says "I" has much more effect than it appears to from observed behaviour.
Yep. Anthropomorphizing humans is a disasterously wrong thing to do. Too bad everyone does it.
No, they just look like they're doing it; saying humans are athropomorphizing would attribute more intentionality to humans than is justified by the data.
Well, the mind seems to. I'm using "mind" here to mean the bit that says "I" and could reflect on itself it if it bothered to and thinks it runs the show and comes up with rationalisations for whatever it does. Listening to these rationalisations, promises, etc. as anything other than vague pointers to behaviour is exceedingly foolish. Occasionally you can encourage the person to use their "mind" less annoyingly.
I think they anthropomorphise as some sort of default reflex. Possibly somewhere halfway down the spinal cord, certainly not around the cerebrum.
I may be wrong, but I think that SilasBarta is pointing out, maybe with some tongue-in-cheek, that you can't accuse humans of anthropomorphizing other humans without yourself being guilty of anthropomorphizing those humans whom you accuse.
Edit: Looks like this was the intended reading.
I am finding benefits from trying not to anthropomorphise myself. That is, rather than thinking of my mind as being in control of my actions, I think of myself as a blob of lard which behaves in certain ways. This has actually been a more useful model, so that my mind (which appears to be involved in typing this, though I am quite ready to be persuaded otherwise) can get the things it thinks it wants to happen happening.
I was joking. :-P
Ha ha only serious ;-p
Is this an example? I've been working on paying attention to intention. If I know someone cares about me, but is expressing it poorly, I try to focus on their intent rather than their expression of that intent.
I'd watch their behaviour, which I would also have classed as expression of the intent. Do they show they care? That being the thing you actually want.
If you're wondering why I'm afraid of philosophy, look no further than the fact that this discussion is assigning salience to LW posts in a completely different way to I do.
I mean, it seems to me that where I think an LW post is important and interesting in proportion to how much it helps construct a Friendly AI, how much it gets people to participate in the human project, or the amount of confusion that it permanently and completely dissipates, all of this here is prioritizing LW posts to the extent that they happen to imply positions on famous ongoing philosophical arguments.
That's why I'm afraid to be put into any philosophical tradition, Quinean or otherwise - and why I think I'm justified in saying that their cognitive workflow is not like unto my cognitive workflow.
With this comment at least, you aren't addressing the list of 20+ useful contributions of mainstream philosophy I gave.
Almost none of the items I listed have to do with famous old "problems" like free will or reductionism.
Instead, they're stuff that (1) you're already making direct use of in building FAI, like reflective equilibrium, or (2) stuff that is almost identical to the 'coping with cognitive biases' stuff you've written about so much, like Bishop & Trout (2004), or (3) stuff that is dissolving traditional debates into the cognitive algorithms that produce them, which you seem to think is the defining hallmark of LW-style philosophy, or (4) generally useful stuff like the work on catastrophic risks coming out of FHI at Oxford.
I hope you aren't going to keep insisting that mainstream philosophy has nothing useful to offer after reading my list. On this point, it may be time for you to just say "oops" and move on.
After all, we already agree on most of the important points, like you said. We agree that philosophy is an incredibly diseased discipline. We agree that people shouldn't go out and read Quine. We agree that almost everyone should be reading statistics and AI and cognitive science, not mainstream philosophy. We agree that Eliezer Yudkowsky should not read mainstream philosophy. We agree that "their" cognitive workflow is "not like unto" your cognitive workflow.
So I don't understand why you would continue to insist that nothing (or almost nothing) useful comes out of mainstream philosophy, after the long list of useful things I've provided, many of which you are already using yourself, and many more of which closely parallel what you've been doing on Less Wrong all along, like dissolving traditional debates into cognitive algorithms and examining how to get at the truth more often through awareness and counteracting of our cognitive biases.
The sky won't fall if you admit that some of mainstream philosophy is useful, and that you already make use of some of it. I'm not going to go around recommending people join philosophy programs. This is simply about making use of the resources that are out there. Most of those resources are in statistics and AI and cognitive science and physics and so on. But a very little of it happens to come out of mainstream philosophy, especially from that corner of mainstream philosophy called Quinean naturalism which shares lots of (basic) assumptions with Less Wrong philosophy.
As you know, this stuff matters. We're trying to save the world, here. Either some useful stuff comes out of mainstream philosophy, or it doesn't. There is a correct answer to that question. And the correct answer is that some useful stuff does come out of mainstream philosophy - as you well know, because you're already making use of it.
I think it would be good for LessWrong to have a bit more academic philosophers and students of philosophy, to have a slightly higher philosophers/programmers ratio (as long as it doesn't come with the expectation that everybody should understand a lot of concepts in philosophy that aren't in the sequences).
You still haven't given an actual use case for your sense of "useful", only historical priority (the qualifier "come out" is telling, for example), and haven't connected your discussion that involves the word "useful" to the use case Eliezer assumes (even where you answered that side of the discussion without using the word, by agreeing that particular use cases for mainstream philosophy are a loss). It's an argument about definition of "useful", or something hiding behind this equivocation.
I suggest tabooing "useful", when applied to literature (as opposed to activity with stated purpose) on your side.
Eliezer and I, over the course of our long discussion, have come to some understanding of what would constitute useful. Though, Philosophy_Tutor suggested that Eliezer taboo his sense of "useful" before trying to declare every item on my list as useless.
Whether or not I can provide a set of necessary and sufficient conditions for "useful", I've repeatedly pointed out that:
Several works from mainstream philosophy do the same things he has spent a great deal of time doing and advocating on Less Wrong, so if he thinks those works are useless then it would appear he thinks much of what he has done on Less Wrong is uesless.
Quite a few works from mainstream philosophy have been used by him, so presumably he finds them useful.
I can't believe how difficult it is to convince some people that some useful things come out of mainstream philosophy. To me, it's a trivial point. Those resisting this truth keep trying to change the subject and make it about how philosophy is a diseased subject (agreed!), how we shouldn't read Quine (agreed!), how other subjects are more important and useful (agreed!), and so on.
I agree that you've agreed on many specific things. I suggest that the sense of remaining disagreement is currently confused through refusing to taboo "useful". You use one definition, he uses a different one, and there is possibly genuine disagreement in there somewhere, but you won't be able to find it without again switching to more specific discussion.
Also, taboo doesn't work by giving a definition, instead you explain whatever you wanted without using the concept explicitly (so it's always a definition in a specific context).
For example:
Instead of debating this point of the definition (and what constitutes "being used"), consider the questions of whether Eliezer agrees that he was influenced (in any sense) by quite a few works from mainstream philosophy (obviously), whether they provided insights that would've been unavailable otherwise (probably not), whether they happen to already contain some of the same basic insights found elsewhere (yes), whether they originate them (it depends), etc.
It's a long list, not as satisfying as the simple "useful/not", but this is the way to unpack the disagreement. And even if you agree on every fact, his sense of "useful" can disagree with yours.
I'll wait to see if Eliezer really thinks we aren't on the same page about the meaning of 'useful'.
If reflective equilibrium, which plays a central role in Eliezer's plan (CEV) to save humanity, isn't useful, then I will be very surprised, and we will seem to be using different definitions of the term "useful."
Has he repudiated the usefulness of reflective equilibrium (or of the concept, or the term)? I recall that he's used it himself in some of the more recent summaries of CEV.
If it's not immediately obvious how an argument connects to a specific implementable policy or empirical fact, default is to covertly interpret it as being about status.
Since there are both good and bad things about philosophy, we can choose to emphasize the good (which accords philosophers and those who read them higher status) or emphasize the bad (which accords people who do their own work and ignore mainstream philosophy higher status).
If there are no consequences to this choice, it's more pleasant to dwell upon the bad: after all, the worse mainstream philosophy does, the more useful and original this makes our community; the better mainstream philosophy does, the more it suggests our community is a relatively minor phenomenon within a broader movement of other people with more resources and prestige than ourselves (and the more those of us whose time is worth less than Eliezer's should be reading philosophy journals instead of doing something less mind-numbing).
I think this community is smart enough to avoid many such biases if given a real question with a truth-value, but given a vague open question like "Yay philosophy - yes or no?" of course we're going to take the side that makes us feel better.
I think the solution is to present specific insights of Quinean philosophy in more depth, which you already seem like you're planning to do.
Maybe my original post gave the wrong impression of "which side I'm on." (Yay philosophy or no?) Like Quine and Yudkowsky, I've generally considered myself an "anti-philosophy philosopher."
But you're right that such vague questions and categorizations are not really the point. The solution is to present specific useful insights of mainstream philosophy, and let the LW community make use of them. I've done that in brief, here, and am working on posts to elaborate some of those items in more detail.
What disappoints me is the double standard being used (by some) for what counts as "useful" when presented in AI books or on Less Wrong, versus what counts as "useful" when it happens to come from mainstream philosophy.
I don't think there is double standard involved.
There are use cases (plans) that distinguish LW from mainstream philosophy that make philosophy less useful for those plans. There are other use cases where philosophy would be more useful. Making an overall judgment would depend on which use cases are important.
The concept of "useful" that leads to a classification which marks philosophy "not useful" might be one you don't endorse, but we already discussed a few examples that show that such concepts can be natural, even if you'd prefer not to identify them with "usefulness".
A double standard would filter evidence differently when considering the things it's double-standard about. If we are talking about particular use cases, I don't think there was significant distortion of attention paid for either case. A point where evidence could be filtered in favor of LW would be focus on particular use cases, but that charge depends on the importance of those use cases and their alternatives to the people selecting them. So far, you didn't give such a selection that favors philosophy, and in fact you've agreed on the status of the use cases named by others.
So, apart from your intuition that "useful" is an applicable label, not much about the rules of reasoning and motivation about your claim was given. Why is it interesting to discuss whether mainstream philosophy is "useful" in the sense you mean this concept? If we are to discuss it, what kinds of arguments would tell us more about this fact? Can you find effective arguments about other people's concepts of usefulness, given that the intuitive appeals made so far failed? How is your choice of concept of "usefulness" related to other people's concepts, apart from the use of the same label? (Words/concepts can be wrong, but to argue that a word is wrong with a person who doesn't see it so would require a more specific argument or reasoning heuristic.)
Since there seems to be no known easy way of making progress on discussing each other's concepts, and the motivation seems to be solely to salve intuition, I think there is no ground for further object-level argument.
I love to read and write interesting things - which is why I take to heart Eliezer's constant warning to be wary of things that are fun to argue.
But interestingness was not the point of my post. Utility to FAI and other Less Wrong projects was the point. My point was that mainstream philosophy sometimes offers things of utility to Less Wrong. And I gave a long list of examples. Some of them are things (from mainstream philosophy) that Eliezer and Less Wrong are already making profitable use of. Others are things that Less Wrong had not mentioned before I arrived, but are doing very much the same sorts of things that Less Wrong values - for example dissolution-to-algorithm and strategies for overcoming biases. Had these things been written up as Less Wrong posts, it seems they'd have been well-received. And in cases where they have been written up as Less Wrong posts, they have been well-received. My continuing discussion in this thread has been to suggest that therefore, some useful things do come from mainstream philosophy, and need not be ignored simply because of the genre or industry they come from.
By "useful" I just mean "possessing utility toward some goal." By "useful to Less Wrong", then, I mean "possessing utility toward a goal of Less Wrong's/Eliezer's." For example, both reflective equilibrium and Epistemology and the Psychology of Human Judgment possess that kind of utility. That's a very rough sketch, anyway.
But no, I don't have time to write up a 30-page conceptual analysis of what it means for something to be "useful."
But I think I still don't understand what you mean. Maybe an example would help. A good one would be this: Is there a sense in which reflective equilibrium (a theory or process that happens to come from mainstream philosophy) is not useful to Eliezer, despite the fact that it plays a central role in CEV, his plan to save humanity from unfriendly AI?
Another one would be this: Is there a sense in which Eliezer's writing on how to be aware of and counteract common cognitive biases is useful, but the nearly identical content in Bishop & Trout's Epistemology and the Psychology of Human Judgment (which happens to come from mainstream philosophy) is not useful?
Sounds plausible, and if true, a useful observation.
"Yay philosophy - yes or no?" and questions of that ilk seem like an interesting question to actually ask people.
You could, for instance, make a debate team lay out the pro and con positions.
A lot of the "nay philosophy" end up doing philosophy, even while they continue to say "nay philosophy". So I have a hard time taking the opinion at face value.
Moreover it's not like there is one kind of thinking, philosophy, and another kind of thinking, non-philosophy. Any kind of evidence or argument could in principle be employed by someone calling himself a philosopher - or, inversely, by someone calling himself a non-philosopher. If you suddenly have a bright idea and start developing it into an essay, I submit that you don't necessarily know whether, once the idea has fully bloomed, it will be considered philosophy or non-philosophy.
I don't know whether it's true that science used to be considered a subtopic of philosophy ("natural philosophy"), but it seems entirely plausible that it was all philosophy but that at some point there was a terminological exodus, when physicists stopped calling themselves philosophers. In that older, more inclusive sense, then anyone who says "nay philosophy" is also saying "nay science". Keeping that in mind, what we now call "philosophy" might instead be called, "what's left of philosophy after the great terminological exodus".
Of course "what's left" is also called "the dregs". In light of that, what we all "philosophy" might instead be called "the dregs of philosophy".
That is exactly true. The old term for what we nowadays call "natural science" was "natural philosophy." There are still relics of this old terminology, most notably that in English the title "doctor of philosophy" (or the Latin version thereof) is still used by physicists and other natural scientists. The "terminological exodus" you refer to happened only in the 19th century.
Isn't the smart move there not to play? What would make that the LW move?
I'm worried part of this debate is just about status. When someone comes in and says "Hey, you guys should really pay more attention to what x group of people with y credentials says about z" it reminds everyone here, most of whom lack y credentials that society doesn't recognize them as an authority on z and so they are some how less valuable than group x. So there is an impulse to say that z is obvious, that z doesn't matter or that having y isn't really a good indicator of being right about z. That way, people here don't lose status relative to group x.
Conversely, members of group x probably put money and effort into getting credential y and will be offended by the suggestion that what they know about doesn't matter, that it is obvious or that their having credential y doesn't indicate they know anything more than anyone else.
Me, I have an undergraduate degree in philosophy which I value so I'm sure I get a little defensive when philosophy is mocked or criticized around here. But most people here probably fit in the first category. Eliezer, being a human being like everybody else, is likely a little insecure about his lack of a formal education and perhaps particularly apt to deny an academic community status as domain experts in a fields he's worked in (even though he is certainly right that formal credentials are overvalued).
I think a lot of this argument isn't really a disagreement over what is valuable and what isn't- it's just people emphasizing or de-emphasizing different ideas and writers to make themselves look higher status.
...
These statements have no content they just say "My stuff is better than your stuff".
I think such debates unavoidably include status motivations. We are status-oriented, signaling creatures. Politics mattered in our ancestral environment.
Of course you know that I never said anything like either of the parody quotes provided. And I'm not trying to stay Quinean philosophy is better than Less Wrong. The claim I'm making is a very weak claim: that some useful stuff comes out of mainstream philosophy, and Less Wrong shouldn't ignore it when that happens just because the source happens to be mainstream philosophy.
Yes. But you're right so that side had to be a strawman, didn't it?
I'm sorry; what do you mean?
Are you, in your view, having The Problem with Non-Philosophers again?
I'm late, but… is there substantial chain of cause and effect between the discovery of useful conclusions from mainstream philosophy, and the use of those conclusions by Eliezer? Counter-factually, if those conclusions were not drawn, would it be less likely that Eliezer found them anyway?
Eliezer seems to deny this chain of cause and effect. I wonder to what extent you think such a denial is unjustified.
It seems to me that the disagreement might be over the adjective "mainstream". To me, that connotes what's being mentioned (not covered in detail, merely mentioned) in broad overviews such as freshman introductory classes or non-major classes at college. As an analogy, in physics both general relativity and quantum mechanics are mainstream. They get mentioned in these contexts, though not, of course, covered. Something like timeless physics does not.
How much of the standard philosophy curriculum covers Quinean Naturalism?
I dunno, I think Eliezer and I are clear on what mainstream philosophy is. And if anything is mainstream, it's John Rawls and Oxford University professors whose work Eliezer is already making use of.
Well, when I see:
That does not make me think that "mainstream philosophy" as a whole is doing useful work. Localized individuals and small strains appear to be. But even when the small strains are taken seriously in mainstream philosophy, that's not the same as mainstream philosophy doing said work, and labeling any advances as "here's mainstream philosophy doing good work" seems to be misleading.
No, mainstream philosophy "as a whole" is not doing useful work. That's what the central section of my original post was about: Non-Quinean philosophy, and how its entire method is fundamentally flawed.
Even quite a lot of Quinean naturalistic philosophy is not doing useful work.
I'm not trying to mislead anybody. But Eliezer has apparently taken the extreme position that mainstream philosophy in general is worthless, so I made a long list of useful things that have come from mainstream philosophy - and some of it is not even from the most productive strain of mainstream philosophy, what I've been calling "Quinean naturalism." Useful things sometimes come from unexpected sources.
In the above quote the following replacements have been made. philosophy -> religion Quinean -> Christian
There are many ideas from religion that are not useless. It is not often the most productive source to learn from either however. Why filter ideas from religion texts when better sources are available or when it is easier to recreate them in within in a better framework; a framework that actual justifies the idea. This is also important because in my experience people fail to filter constantly and end up accepting bad ideas.
I do not see EY arguing that main stream philosophy has not useful nuggets. I seem him arguing that filtering for those nugets in general makes the process too costly. I see you arguing that "Quinean naturalism" is a rich vien of philosophy and worth mining for nuggets. If you want to prove the worth of mine "Quinean naturalism" you have to display nuggets that EY has not found through better means already.
I did list such nuggets that EY has not found through other means already, including several instances of "dissolution-to-algorithm", which EY seems to think of as the hallmark of LW-style philosophy.
I wouldn't call mainstream philosophy a "rich vein" that is (for most people) worth mining for nuggets. I've specifically said that people will get far more value reading statistics and AI and cognitive science. I've specifically said that EY should not be mining mainstream philosophy. What I'm saying is that if useful stuff happens to come from mainstream philosophy, why ignore it? It's people like myself who are already familiar with mainstream philosophy, and for whom it doesn't take much effort to list 20+ useful contributions of mainstream philosophy, who should bring those useful nuggets to the attention of Less Wrong.
What seems strange to me is to draw an arbitrary boundary around mainstream philosophy and say, "If it happens to come from here, we don't want it." And I think Eliezer already agrees with this, since of course he is already making use of several things from mainstream philosophy. But on the other hand, he seems to be insisting that mainstream philosophy has nothing (or almost nothing) useful to offer.
I don't understand the distinction you're making. These two statements mean the exact same thing to me: in general, mainstream philosophy is useless, though exceptions exist.
Admittedly. That's not a good reason to look there, until the expected sources are exhausted.
What I'm trying to say is that the vast majority of mainstream philosophy is useless, but some of it is useful, and I gave examples.
I've also repeatedly agreed that most people should not be reading mainstream philosophy. Much better to learn statistics and AI and cognitive science. But for those already familiar with philosophy, for whom it's not that difficult to name 20 useful ideas from mainstream philosophy, then... why not make use of them? It makes no sense to draw an arbitrary boundary around mainstream philosophy and say "If it comes from here, I don't want it." That's silly.
I've frequently been criticized for suggesting that you hold that attitude. The usual response is that LW is not about friendly AI or has not much to do with the SIAI.