Posts
Comments
Reaching?
This one felt quite LW-relevant:
If $1 million makes you happy, that doesn't mean $10 million will make you 10 times as happy.
It's good to be reminded now and then that dollars are not, in fact, utilons.
The lack of an automatic repair mechanism makes things hairier, but while frozen, the radiation damage will be localized to the cells that get hit by radiation. By the time you get the tech to revive people from cryonic freezing, you'll most likely have the tech to fix/remove/replace the individual damaged cells before reviving someone. I think you're right that radiation won't be a big limiting factor, though it may be an annoying obstacle.
I don't know so much about C-14, but wouldn't potassium decay's effects be small on timescales ~10,000 years? The radioactive natural isotope K-40 has a ridiculously long half life (1.25 billion years, which is why potassium-argon dating is popular for dating really old things) and only composes 0.012% of natural potassium. Potassium's also much less abundant in the body than carbon - only about 140g of a 70kg person is potassium, although admittedly it might be more concentrated in the brain, which is the important part.
ETA - I did calculations, and maybe there is a problem. Suppose 0.012% of K is K-40 by mass. Then I get 0.0168 grams of K-40 in a body, which comes out as 0.00042 moles, 2.53e20 K-40 atoms. With a 1.25 billion year half life that makes 1.40e15 decays after 10,000 years. In absolute terms that's a lot of emitted electrons and positrons. I don't know whether the absolute number (huge) or the relative number (miniscule) is more important though.
This is a nice example of the slipperiness I sometimes notice when I think about how one could test an ev. psych hypothesis. My first thought after reading your comment was 'but won't all those factors be just as correlated with unhappiness and depression as with genetic fitness? Surely there's a less complex explanation here: unhappy people don't like living as much, so they try killing themselves more.'
Then I thought a little more and realized that could also have an ev. psych basis: maybe we evolved to kill ourselves more when we're unhappy with life. But that's a different ev. psych argument than 'we evolved to kill ourselves more when we have low genetic fitness.' Or is it? Maybe we evolved to kill ourselves more when we're unhappy because unhappiness correlates with low genetic fitness? What would it even mean in concrete terms for all of these possible evolutionary explanations to be false?
'Why would I need to demand evidence? My wife freely gives me evidence of her love, all the time!'
Maybe the salesman mostly sells temporary life insurance, and just means that no clients had died while covered?
I'm sorry, this is just an open thread comment, not a top-level post. Aren't we allowed to just chat and get feedback without thoroughly contemplating a subject?
There's nobody compelling you to reply to the comments you feel are too thoroughly contemplating a subject.
Who's right? Who knows. It's a fine opportunity to remain skeptical.
Bullshit. The 'skeptical' thing to do would be to take 30 seconds to think about the theory's physical plausibility before posting it on one's blog, not regurgitate the theory and cover one's ass with an I'm-so-balanced-look-there's-two-sides-to-the-issue fallacy.
TV-frequency EM radiation is non-ionizing, so how's it going to transfer enough energy to your cells to cause cancer? It could heat you up, or it could induce currents within your body. But however much heating it causes, the temperature increase caused by heat insulation from your mattress and cover is surely much greater, and I reckon you'd get stronger induced currents from your alarm clock/computer/ceiling light/bedside lamp or whatever other circuitry's switched on in your bedroom. (And wouldn't you get a weird arrhythmia kicking off before cancer anyway?)
(As long as I'm venting, it's at least a little silly for Kottke to say he's posting it because it's 'interesting' and not because it's 'right,' because surely it's only interesting because it might be right? Bleh.)
Does the lack of a response from EY imply that he's not interested in that sort of change and, if so, is it EY who would be the one to make the decision?
I wouldn't read anything into the lack of response, EY often doesn't comment on meta-discussion. In fact I'd guess there's a good chance he hasn't even seen this thread!
I guess it might be worth raising this in the Spring 2010 meta-thread? Come to think of it, it's been 4+ months since that meta thread was started - it may even be worth someone posting a Summer 2010 meta-thread with this as a topic starter.
I had a houseguest for a few days recently, a long-time reader who has only written a handful of comments, and I commented to him that the quality of discussion on LW is worse than it has ever been, and his reply was, "Well, yeah if you are talking about WrongBot."
I think your houseguest might not have read a representative selection of LW posts; their assessment doesn't ring true for me. I haven't read WrongBot's top-level posts closely (nothing personal - the evolutionary psychology stuff just isn't that interesting to me), but I've skimmed through the resulting threads/comments on them as they've passed through Recent Comments, and they honestly don't look all that bad.
I can think of a few recent posts/discussion topics that I am fairly confident have lower quality than WrongBot's:
- '(One reason) why capitalism is much maligned'
- Daniel_Burfoot's quite rambling series of posts that uses 7000 words just to talk up data compression as an add-on to the scientific method
- whpearson's bit of evolutionary psychology 'Summer vs Winter Strategies'
- MBlume's link to 'Jinnetic Engineering' - the content is good, but it's not meaty enough for a top-level post IMO
- the string of posts a while back dancing around the Sleeping Beauty puzzle and what it meant - there was a lot of good in them, and their comments, but the discussions got really flabby really fast
Enthusiastically seconded.
The only change I'd make is to hide editorial comments when the post leaves editing (instead of deleting them), with a toggle option for logged-in users to carry on viewing them.
Unfortunately, most of the busy smart people only looked at the posts after editing, while the trolls and people with too much free time managed the edit queue, eventually destroying the quality of the site and driving the good users away. It might be possible to salvage that model somehow, though.
I think it is. There are several tricks we could use to give busy-smart people more of a chance to edit posts.
On Kuro5hin, if I remember right, posts left the editing queue automatically after 24 hours, either getting posted or kicked into the bit bucket. Also, users could vote to push the story out of the queue early. If Less Wrong reimplemented this system, we could raise the threshold for voting a story out of editing early, or remove the option entirely. We could even lengthen the period it spends in the editing stage. (This would also have the advantage of filtering out impatient people who couldn't wait 3 days or whatever for their story to post.)
LW's also just got a much smaller troll ratio than Kuro5hin did, which would help a lot.
Upvoted for raising the topic, but the approach I'd prefer is jimrandomh's suggestion of having all posts pass through an editorial stage before being posted 'for real.'
I'm open to suggestions for how I might improve the introduction to the article to make the article more palatable.
I was going to suggest this, but I see you've already added it: thanks for editing in your definition of capitalism at the top of the post. When I first read the post, that was something I thought would improve it. Like SilasBarta I thought it was a bad idea to leave unclear what you were counting as capitalism.
I think that's a clever idea that deserves more eyeballs.
Done. I'm looking forward to either Nancy's substantive reply and apology, or your concession that the issue might be a bit more complicated.
It seems to me that the issue's already been complicated because you've already replied to Nancy impolitely. Now that's happened, it is not really realistic to expect a substantive reply and apology from her simply because you (I, if we're being pedantic) rephrased some of your original remarks more tactfully.
Okay, but the part Nancy ignored when she replied bore directly on (and obviated!) her comment, so she shouldn't have replied to begin with if that was all she had to say. The general point of yours (which I agree with) about the impossibility of replying to everything, doesn't apply.
OK; it sounds like I misinterpreted your earlier comment about 'people complain that ...' as being directed at me, but based on your reply it sounds like it isn't. In which case feel free to disregard the last paragraph of my grandparent comment.
The rudeness is in how she completely ignores the explanation I just gave in the parent comment, of why wide feet would lead to people being prejudiced against you, which obviates her question.
There's no explicit question in the comment of NL's I think you're thinking of, so I imagine you mean that the statements in her comment could be read as implying an already-answered question, which makes the comment rude. That hardly registers on my rudeness detector; unless it's part of a systematic pattern of behavior, it's innocuous IMO.
Still, let me pretend I'm SilasBarta and suppose her comment is rude.
So:
1) I explain why having wide feet leads to people being prejudiced against me.
2) Nancy replies, while ignoring the entire explanation I just gave.
3) [Insert comment I should have made instead of the one I did, which would point out how Nancy just ignored the explanation I gave, but which you don't characterize as obnoxious]
OK, I'm SilasBarta. Nancy's replied to me. Most of my comment seems to have gone right past her and she's replied without having understood me. That means I have failed to make myself as clear to her as I'd like, and I want to fix that. It's her first reply to me, she's not being overtly confrontational, and people often write sloppily when replying to others on the Internet, so let's assume good faith. As such, I reply to emphasize my more detailed explanation of how people with wide feet suffer prejudice, this time without any snitty rhetorical questions (or accusations of bad faith). I might write something like: 'Let me clarify. Although people with wide feet may not suffer much direct prejudice, they nonetheless suffer effective prejudice indirectly because it hurts my ability to signal via e.g. choice of shoes, as I pointed out in my earlier comment.'
The reason I belabor the point is that this issue comes up quite frequently, where people complain that "Yeah, Silas, you had a good point, but goshdarnit, the way you said it gives me sufficient pretense to ignore it wholesale and join the anti-Silas's point bandwagon", and I want someone to finally put their neck out and show me what comment would be an appropriate one to protest the (rude) ignoring of part of my comment when someone replies to it.
I don't believe I did use the way you said what you said as a pretense for ignoring its good points. I do think you might have been right when you tried picking out the pity-oriented subtext of NL's original post, but just because I didn't mention it doesn't mean I ignored it wholesale - it just means I didn't have anything to say in response to it. There are a lot of comments on Less Wrong that make good points - presented abrasively or otherwise - that I don't reply to. (Also, I wouldn't even have complained to you if you hadn't solicited feedback on why people had voted down your original run of comments.)
I don't think I'm capable of answering that question, since I'm not seeing the 'rudeness' in the parent comment posted by Nancy to which your linked comment replies. At any rate, I didn't find that particular comment of yours obnoxious except for the 'pity party' snark, which I basically just wrote off as your usual level of prickliness.
Did you find it obnoxious when Nancy outright ignored the part of the comment where I explained why having wide feet would lead to others being prejudiced against you? Or just the fact of me mentioning this ignoranc ... er, "act of ignoring".
Neither. I found the manner in which you mentioned it obnoxious, not the mention qua mention.
This is what always gets me: no one cares when someone doesn't read a comment and yet still replies to it -- well, to a version of it. Yet when someone points out the rudeness of doing so -- well, then that person's just a terrorist!
You are mistaken. I'm not objecting to your pointing out that NL didn't acknowledge your comment as you wanted her to. I'm objecting to the claim that she replied with a 'pretense of ignorance.'
What gives? If you're going to criticize just one of those two, which one has priority?
The one that employs immoderate hyperbole and launches an ill-grounded accusation of 'pretense' at someone else.
Not one of the downvoters, but the tone of these paragraphs was so overcooked I did consider it for a couple seconds:
And frankly, when the asymmetic bra issue came up, I got pretty scared. Some of the commenters -- and I'm not going to single anyone out -- sound like really angry people in general and I fear that being around them would make their rage spill on to me.
They have this entitlement mentality, where everyone has to make clothes that they like. I think it's what motivates a lot of the crime against retailers.
I mean, how dare they make clothes for other people, right?
Those words and your presumptuous 'are you going to take back your pretense of ignorance about shoe prejudice?' question came across to me as quite obnoxious.
Having thought about it a little longer and updated based on your evidently broader knowledge of bras, my original guess for why the market failure exists does seem pretty unlikely.
what interpretation of the word "probability" does allow you to think that the probability of something is 1 and then change to something other than 1?
Any interpretation where you can fix a broken model. I can imagine a conversation like this...
Prankster: I'm holding a die behind my back. If I roll it, what probability would you assign to a 1 coming up?
cupholder: Is it loaded?
Prankster: No.
cupholder: Are you throwing it in a funny way, like in one of those machines that throws it so it's really likely to come up a 6 or something?
Prankster: No, no funny tricks here. Just rolling it normally.
cupholder: Then you've got a 1/6 probability of rolling a 1.
Prankster: And what about rolling a 2?
cupholder: Well, the same.
Prankster: And so on for all the other numbers, right?
cupholder: Sure.
Prankster: So you assign a probability of 1 to a number between 1 and 6 coming up?
cupholder: Yeah.
Prankster: Surprise! It's 20-sided!
cupholder: Huh. I'd better change my estimate from 1 to 6/20.
You're right, I should've thought of that. I expect it's easier (maybe therefore cheaper?) to manufacture little silicone blobs or whatever than a half-bra, which must partly be why there's a market for the first and not the second.
There's an even more compelling market: women who have had a single mastectomy. I'd be surprised if there weren't medical half-bras out there already for them.
These results seem to support that, though there have also been contradictory reports from people saying that the very aggressiveness was what made them actually think.
Presumably there's heterogeneity in people's reactions to aggressiveness and to soft approaches. Most likely a minority of people react better to aggressive approaches and most people react better to being fed opposing arguments in a sandwich with self-affirmation bread.
The negotiation of where LW threads should be on the 4chan-colloquium continuum is something I would let users handle by interacting with each other in discussions, instead of trying to force it to fit the framework of the karma system. I especially think letting people hide their posts from lurkers and other subsets of the Less Wrong userbase could set a bad precedent.
So would it be right to say your objection is based on the expected utility of working cryonics instead of its probability?
For being a cryonics facility? Is there enough evidence to determine if it could've been just a random drive-by?
But if we start with a prior over all possible six-sided dice and do Bayesian updating, we get a different answer that diverges from fairness more and more as the number of throws goes to infinity!
In this example, what information are we Bayesian updating on?
yes
OK. I agree with that insofar as agents having the same prior entails them having the same model.
aaahhh.... I changed the language of that sentence at least three times before settling on what you saw. Here's what I probably should have posted (and what I was going to post until the last minute):
There's no model checking because there is only one model - the correct model.
That is probably intuitively easier to grasp, but I think a bit inconsistent with my language in the rest of the post. The language is somewhat difficult here because our uncertainty is simultaneously a map and a territory.
Ah, I think I get you; a PB (perfect Bayesian) doesn't see a need to test their model because whatever specific proposition they're investigating implies a particular correct model.
For the record, I thought this sentence was perfectly clear. But I am a statistics grad student, so don't consider me representative.
Yeah, I figured you wouldn't have trouble with it since you talked about taking classes in this stuff - that footnote was intended for any lurkers who might be reading this. (I expected quite a few lurkers to be reading this given how often the Gelman and Shalizi paper's been linked here.)
Are you asserting that this a catch for my position? Or the "never look back" approach to priors? What you are saying seems to support my argument.
It's a catch for the latter, the PB. In reality most scientists typically don't have a wholly unambiguous proposition worked out that they're testing - or the proposition they are testing is actually not a good representation of the real situation.
My implicit definition of perfect Bayesian is characterized by these propostions:
- There is a correct prior probability (as in, before you see any evidence, e.g. occam priors) for every proposition
- Given a particular set of evidence, there is a correct posterior probability for any proposition
OK, this is interesting: I think our ideas of perfect Bayesians might be quite different. I agree that #1 is part of how a perfect Bayesian thinks, if by 'a correct prior...before you see any evidence' you have the maximum entropy prior in mind.
I'm less sure what 'correct posterior' means in #2. Am I right to interpret it as saying that given a prior and a particular set of evidence for some empirical question, all perfect Bayesians should get the same posterior probability distribution after updating the prior with the evidence?
If we knew exactly what our priors were and how to exactly calculate our posteriors, then your steps 1-6 is exactly how we should operate. There's no model checking because there is no model.
There has to be a model because the model is what we use to calculate likelihoods.
The rationale for model checking should be pretty clear ...
Agree with this whole paragraph. I am in favor of model checking; my beef is with (what I understand to be) Perfect Bayesianism, which doesn't seem to include a step for stepping outside the current model and checking that the model itself - and not just the parameter values - makes sense in light of new data.
I spent a few days wondering how it squared with the Baysian philosophy of induction, and then what I took to be obvious answer came to me (while discussing it with my professor actually): we're modeling our uncertainty.
The catch here (if I'm interpreting Gelman and Shalizi correctly) is that building a sub-model of our uncertainty into our model isn't good enough if that sub-model gets blindsided with unmodeled uncertainty that can't be accounted for just by juggling probability density around in our parameter space.* From page 8 of their preprint:
If nothing else, our own experience suggests that however many different specifications we think of, there are always others which had not occurred to us, but cannot be immediately dismissed a priori, if only because they can be seen as alternative approximations to the ones we made. Yet the Bayesian agent is required to start with a prior distribution whose support covers all alternatives that could be considered.
* This must be one of the most dense/opaque sentences I've posted on Less Wrong. If anyone cares enough about this comment to want me to try and break down what it means with an example, I can give that a shot.
I'm not sure about this. It's most likely that anything your kid does in life will get done by someone else instead.
True - we might call the expected utility strangers get a wash because of this substitution effect. If we say the expected value most people get from me having a child is nil, it doesn't contribute to the net expected value, but nor does it make it less positive.
There is also some evidence that having children decreases your happiness (though there may be other reasons to have kids).
It sounds as though that data's based on samples of all types of parents, so it may not have much bearing on the subset of parents who (a) have stable (thanks NL!) high living standards, (b) are good at being parents, and (c) wanted their children. (Of course this just means the evidence is weak, not completely irrelevant.)
But even if this is true, it's still not enough for antinatalism. Increasing total utility is not enough justification to create a life.
That's a good point, I know of nothing in utilitarianism that says whose utility I should care about.
The act of creation makes you responsible for the utility of the individual created, and you have a duty not to create an entity you have reason to think may have negative personal utility. (Strict utilitarians will disagree.)
Whether or not someone agrees with this is going to depend on how much they care about risk aversion in addition to expected utility. (Prediction: antinatalists are more risk averse.) I think my personal level of risk aversion is too low for me to agree that I shouldn't make any entity that has a chance of suffering negative personal utility.
Points taken.
Let me restate what I mean more formally. Conditional on high living standards, high-quality parenting, and desire to raise a child, one can reasonably calculate that the expected utility (to myself, to the potential child and to others) of having the child is higher than the expected utility of not having a child. In which case I wouldn't think the antinatalism position has legs.
After the fact model checking is completely incompatible with perfect Bayesianism, if we define perfect Bayesianism as
- Define a model with some parameters.
- Pick a prior over the parameters.
- Collect evidence.
- Calculate the likelihood using the evidence and model.
- Calculate the posterior by multiplying the prior by the likelihood.
- When new evidence comes in, set the prior to the posterior and go to step 4.
There's no step for checking if you should reject the model; there's no provision here for deciding if you 'just have really wrong priors.' In practice, of course, we often do check to see if the model makes sense in light of new evidence, but then I wouldn't think we're operating like perfect Bayesians any more. I would expect a perfect Bayesian to operate according to the Cox-Jaynes-Yudkowsky way of thinking, which (if I understand them right) has no provision for model checking, only for updating according to the prior (or previous posterior) and likelihood.
Do people here really think that antinatalism is silly?
A data point: I don't think antinatalism (as defined by Roko above - 'it is a bad thing to create people') is silly under every set of circumstances, but neither is it obviously true under all circumstances. If my standard of living is phenomenally awful, and I knew my child's life would be equally bad, it'd be bad to have a child. But if I were living it up, knew I could be a good parent, and wanted a kid, what would be so awful about having one?
A good illustration of multiple discovery (not strictly 'discovery' in this case, but anyway) too:
While Ettinger was the first, most articulate, and most scientifically credible person to argue the idea of cryonics,[citation needed] he was not the only one. In 1962, Evan Cooper had authored a manuscript entitled Immortality, Scientifically, Physically, Now under the pseudonym "N. Durhing".[8] Cooper's book contained the same argument as did Ettinger's, but it lacked both scientific and technical rigor and was not of publication quality.[citation needed]
In the long run, it's all good - I think it's a decent paper, and I suppose this way more eyeballs see it than if I was the only one to post it. (Not to say that we should make a regular habit of linking things four times :-)
My understanding is that historically, schizophrenia has been presumed to have a partly genetic cause since around 1910, out of which grew an intermittent research program of family and twin studies to probe schizophrenia genetics. An opposing camp that emphasized environmental effects emerged in the wake of the Nazi eugenics program and the realization that complex psychological traits needn't follow trivial Mendelian patterns of inheritance. Both research traditions continue to the present day.
Edit to add - Franz Josef Kallman, whose bibliography in schizophrenia genetics I somewhat glibly linked to in the grandparent comment, is one of the scientists who was most firmly in the genetic camp. His work (so far as I know) dominated the study of schizophrenia's causes between the World Wars, and for some time afterwards.
?+schizophrenia)
Agreed. I wish they'd stick to calling hard-to-read graphics like this 'visualizations' - the word 'infographics' implies a graphic designed to efficiently display information.
The worst part is it wouldn't be hard to improve the graphic. They could drop the annoying 84-item list and just directly write the emotions in the 84 slots around the circle instead of using numbers. Enlarge the circle and blow up the font size a bit - then they can put the A to J list of cultures into the empty middle of the circle so you don't have to keep looking off the side to cross-reference it. That'd help, even if it wouldn't fix it.
Edit - I see that when they used that infographic as their book's cover, they gave up on the idea of making it a real infographic and just made it into a pretty flower!
I wonder if this counts as evidence for my heuristic of judging how seriously to take someone's belief on a complicated scientific subject by looking to see if they get the right answer on easier scientific questions.
It might well have been clear from the quote itself, but not to me - I just read the quote as saying Bayesian thinking and Bayesian methods haven't become more popular in science, which doesn't mesh with my intuition/experience.
I don't know of a way to get superscripts with Markdown markup, but if you pull up your Windows Character Map (or your operating system's equivalent), there should be superscript 1 and 2 characters to paste in.
I'd have expected that men would lie about having more partners rather than fewer, but that might be mere stereotyping on my part.
Or maybe the sample's not representative of most men - the sample was of Midwestern psychology undergraduate students.
It's apparently been put to use with some success. Clark Glymour - a philosophy professor who helped develop TETRAD - wrote a long review of The Bell Curve that lists applications of an earlier version of TETRAD (see section 6 of the review):
Several other applications have been made of the techniques, for example:
Spirtes et al. (1993) used published data on a small observational sample of Spartina grass from the Cape Fear estuary to correctly predict - contrary both to regression results and expert opinion - the outcome of an unpublished greenhouse experiment on the influence of salinity, pH and aeration on growth.
Druzdzel and Glymour (1994) used data from the US News and World Report survey of American colleges and universities to predict the effect on dropout rates of manipulating average SAT scores of freshman classes. The prediction was confirmed at Carnegie Mellon University.
Waldemark used the techniques to recalibrate a mass spectrometer aboard a Swedish satellite, reducing errors by half.
Shipley (1995, 1997, in review) used the techniques to model a variety of biological problems, and developed adaptations of them for small sample problems.
Akleman et al. (1997) have found that the graphical model search techniques do as well or better than standard time series regression techniques based on statistical loss functions at out of sample predictions for data on exchange rates and corn prices.
Personally I find it a little odd that such a useful tool is still so obscure, but I guess a lot of scientists are loath to change tools and techniques.
One possible way to get started is to do what the 'Distilling Free-Form Natural Laws from Experimental Data' project did: feed measurements of time and other variables of interest into a computer program which uses a genetic algorithm to build functions that best represent one variable as a function of itself and the other variables. The Science article is paywalled but available elsewhere. (See also this bunch of presentation slides.)
They also have software for you to do this at home.
"Silas, there is no Bayesian ‘revival’ in science. There is one amongst people who wish to reduce science to a mechanical procedure." – Gene Callahan
Am I the only one who finds this extremely unlikely? So far as I know, Bayesian methods have become massively more popular in science over the last 50 years. (Count JSTOR hits for the word 'Bayesian,' for example, and watch the numbers shoot up over time!)
The problem that I hear most often in regard to mechanizing this process has the basic form, "Obviously, you need a human in the loop because of all the cases where you need to be able to recognize that a correlation is spurious, and thus to ignore it, and that comes from having good background knowledge."
Those people should be glad they've never heard of TETRAD - their heads might have exploded!