RolfAndreassen

Karma: 4,799

RolfAndreassen 17 Apr 2009 19:45 UTC
6 points
on: Anti-rationality quotes
I think there is a large asymmetry between rationality quotes and unrationality quotes. The first strive to set a good example, or in some cases just to be funny while still true. So you can go through a long list of quotes and pick out the best ones, and accomplish your purpose. But unrationality quotes are attempting to demonstrate a preponderance, even a dominance, of dis-rationality over rationality in a particular tradition. For this it is not sufficient to cherry-pick a round dozen quotes from 2 millennia; you must also demonstrate that there are not counterweights to the ones you show. For all we know from what you’ve shown, these are the bottom 10% of Christian thinking; the rationality quotes, on the other hand, are avowedly picked from the top 10%.

Your thesis may still be true, indeed I think it is, but the quotes do not demonstrate it.

That said, the post might be a useful Dark Side weapon for making the first chink in someone’s armour of rationalisations.

RolfAndreassen 8 May 2009 19:43 UTC
1 point
in reply to: Cyan’s comment on: A Request for Open Problems
“The uniform distribution centered at c” does not seem to make sense. Did you perchance mean the Gaussian distribution? Further, ‘deviates’ looks like jargon to me. Can we use ‘samples’? I would therefore rephrase as follows, with specific example to hang one’s visualisation on:

Heights of male humans are known to have a Gaussian distribution of width 10 cm around some central value ; unfortunately you have forgotten what the central value is. Joe is 180 cm, Stephen is 170 cm. The probability that is between these two heights is 50%; explain why. Then find a better confidence interval for .

RolfAndreassen 18 May 2009 17:03 UTC
7 points
on: “What Is Wrong With Our Thoughts”
Possibly Stove intended this only as an extended Take That to philosophers he dislikes; but it seems to me that he is a bit too dismissive of his own project, the ‘nosology’. Without wanting a Fully General Counterargument, I think it might be useful to have a set of, say, five or six different classes of erroneous statements; and I also think Stove is too eager to insist on the singularity of each of his examples. For example, he states that the objection “not verifiable” cannot be applied to his example 8; I don’t see why not. Anything whose “name and nature are not to be revealed” has just been declared unverifiable, no? Similarly 3 through 7 look pretty unverifiable to me.

Then he has some examples further down the list which look reasonably testable, such as 13 : “3 is a lucky number”. One could easily do an experiment on this by submitting lottery tickets with and without 3′s filled in; and as for 14, I think a simple “false-to-fact” would suffice to dismiss it.

So far then there are three classifications: False to fact, contradiction, meaningless through having no connection to observation. We may need a fourth to cover such statements as 26: “The tie which unites the number three to its properties (such as primeness) is inexplicable”. This seems somehow vaguely related to observation, in that there does seem to be something called three which has the property of primeness, and nobody has really explained the tie between triples of objects and these properties. (It is perhaps not strongly coupled to observation, but I hesitate to dismiss it completely on that ground.) I suggest a fourth classification of ‘uninteresting’ or ‘unfruitful’: A proposition which, when adopted as an axiom, yields few or no deductions, is unfruitful. One might also call it the ‘So What’ error: Making statements which even if true are not useful to know.

There does seem to be some overlap here; for example, Stove’s 25: “Five is of the same substance as three, co-eternal with three, very three of three: it is only in their attributes that three and five are different.” This looks to me quite unverifiable, but even if it were true, So What? What conclusions or prediction would you draw from this?

Contrary to Stove, I think these four will cover all his list: False to fact, contradiction, meaningless, and So What. I am not certain, however, whether this insight is useful.

RolfAndreassen 27 May 2009 15:57 UTC
0 points
in reply to: cousin_it’s comment on: “What Is Wrong With Our Thoughts”
Again, it does seem observable that nobody has explained why three is prime and four isn’t. (I’m not sure you can actually use ‘why’ in an intelligible way here; possibly I’m being confused by non-mathematical language applied to math.) It’s not an observation I would expect anyone to care about, and possibly it may be the equivalent of nobody having seen something invisible; but it does seem to make a statement that could in principle have gone the other way.

RolfAndreassen 9 Jun 2009 18:13 UTC
1 point
on: The Aumann’s agreement theorem game (guess 2/3 of the average)
I think there may be some signaling issues to be considered here. We all presumably self-identify to some degree as rationalists, and want to validate Less Wrong as a rationalist community. The more rational a community is, the closer to zero the average guess ought to come. So by guessing zero, lowering the average, you contribute to a signal that Less Wrong is a rational sort of place, and validate your own participation. This could be avoided by not revealing the average guess, but then we’ve gone away from interesting sociological experiment and into forum games.

RolfAndreassen 16 Jun 2009 19:52 UTC
4 points
on: Intelligence enhancement as existential risk mitigation
I do not know if this strategy would apply in the Western world; but in Africa, I think much could be gained simply by nutritional intervention. IQ is, to a good approximation, 50% genetic; much of the environmental effect is childhood nutrition; it follows that widespread distribution of vitamins might have a nice effect over time, in addition to the more usual benefits. It also seems possible that this might work in those strata of the American population that subsist mainly on fast food, although I expect the effect would be less—likely there are already patchwork government programs that distribute vitamins to the poor.

RolfAndreassen 5 Jul 2009 18:48 UTC
5 points
in reply to: RobinHanson’s comment on: Rationality Quotes—July 2009
The conversation took place in 1965; if Feynman believed that, as is likely, he was very probably correct. On the other hand, a lot of people probably thought they knew something about psychology; it was a popular subject at the time.

RolfAndreassen 5 Jul 2009 19:05 UTC
0 points
in reply to: Eliezer Yudkowsky’s comment on: Rationality Quotes—July 2009
Although that is true, I think people who believe in freeplegrunge also believe that freeplegrunge implies immortality; indeed, that’s the reason they believe in freeplegrunge, although they rarely admit it. So your statement misses the mark a bit; it’s not the lack of freeplegrunge people object to, it’s the lack of immortality. And that is not only depressing, it really is a property of particles as opposed to freeplegrunge; at least this is true in the minds of freeplegrungists, because of the very strong coupling between their freeplegrunge-belief and their immortality-belief.

RolfAndreassen 7 Jul 2009 19:20 UTC
6 points
in reply to: CronoDAS’s comment on: Can chess be a game of luck?
Perhaps it would be useful to make a distinction between the game of chess, which is mostly skill, and the game of betting on chess, which is mostly luck with fairly well-known probabilities. It looks to me as though the original article is conflating the two very badly, and that this is the cause of much confusion.

RolfAndreassen 17 Jul 2009 17:59 UTC
0 points
in reply to: timtyler’s comment on: The Popularization Bias
By the time your civilisation is taking advantage of black holes, it’s large enough that even a small temperature difference can scale to quite a bit of negentropy. Further, you don’t have to be in orbit, you can build a Dyson shell around the hole at such a distance that the surface gravity is one g. (Or several shells, if people prefer different levels of gravity.) Then there’s no orbital velocity to deal with. (And in any case, you could brake by tidal friction and extract some entropy that way.) Or to be shorter, why are you objecting to the practical details of a thought experiment? Nothing about the game theory relies on black holes or the particular exponent 2; it could just as well be mass^1.5, and the analysis would remain the same although the numbers would change a bit.

RolfAndreassen 19 Jul 2009 21:16 UTC
2 points
in reply to: billswift’s comment on: The Popularization Bias
So you have to expend a bit of energy moving it back to the midpoint every so often. What are attitude jets for?

RolfAndreassen 26 Jul 2009 18:39 UTC
2 points
on: Bayesian Flame
Perhaps we can try an experiment? We have here, apparently, both Bayesians and frequentists; or at a minimum, people knowledgeable enough to be able to apply both methods. Suppose I generate 25 data points from some distribution whose nature I do not disclose, and ask for estimates of the true mean and standard deviation, from a Bayesian and a frequentist? The underlying analysis would also be welcome. If necessary we could extend this to 100 sets of data points, ask for 95% confidence intervals, and see if the methods are well calibrated. (This does probably require some better method of transferring data than blog comments, though.)

As a start, here is one data set:

617.91 16.8539 83.4021 141.504 545.112 215.863 553.168 414.435 4.71129 609.623 117.189 −102.648 647.449 283.57 286.838 710.811 505.826 79.3366 171.816 105.332 540.313 429.298 −314.32 255.93 382.471

It is possible that this task does not have sufficient difficulty to distinguish between the approaches. If so, how can we add constraints to get different answers?

RolfAndreassen 26 Jul 2009 22:07 UTC
6 points
on: Bayesian Flame
I had another thought on the subject. Consider flipping a coin; a Bayesian says that the 50% estimate of getting tails is just your own inability to predict with sufficient accuracy; a frequentist says that the 50% is a property of the coin—or to be less straw-making about it, a property of large sets of indistinguishable coin-flips. So, ok, in principle you could build a coin-predictor and remove the uncertainty. But now consider an electron passing through a beam splitter. Here there is no method even in principle of predicting which Everett branch you find yourself in. (Given some reasonable assumptions about locality and such.) The coin has hidden variables like the precise location of your thumb and the exact force your muscles apply to it; if you were smart enough, you could tease a prediction out of them. But an electron has no such hidden properties. Is it not reasonable, then, to say that the 50% chance really is a property of the electron, and not the predictor?

RolfAndreassen 27 Jul 2009 23:44 UTC
0 points
in reply to: GuySrinivasan’s comment on: Bayesian Flame
Ok, but I don’t think that matters for the question of frequentist versus Bayesian. You’re still saying that the 50% is a property of something other than your own uncertainty.

Moving the problem to lexical uncertainty seems to me to rely on moving the question in time; you can only do this after you’ve done the experiment but before you’ve looked at the measurement. This feels to me like asking a different question.

RolfAndreassen 28 Jul 2009 16:37 UTC
3 points
in reply to: marks’s comment on: Bayesian Flame
Hmm. I don’t know if I’m a very random source of distributions; humans are notoriously bad at randomness, and there are only so many distributions readily available in standard libraries. But in any case, I don’t see this as a difficulty; a real-world problem is under no obligation to give you an easily recognised distribution. If Bayesians do better when the distribution is unknown, good for them. And if not, tough beans. That is precisely the sort of thing we’re trying to measure!

I don’t think, though, that the existence of a Bayesian who can win, based on knowing what distributions I’m likely to use, is a very strong statement. Similarly there exists a frequentist who can win based on watching over my shoulder when I wrote the program! You can always win by invoking special knowledge. This does not say anything about what would happen in a real-world problem, where special knowledge is not available.

RolfAndreassen 29 Jul 2009 17:41 UTC
0 points
in reply to: marks’s comment on: Bayesian Flame
The small samples are a constraint imposed by the limits of blog comments; there’s a limit to how many numbers I would feel comfortable spamming this place with. If we got some volunteers, we might do a more serious sample size using hosted ROOT ntuples or zipping up some plain ASCII.

I do know how to sample from arbitrary distributions; I should have specified that the space of distributions is those for which I don’t have to think for more than a minute or so, or in other words, someone has already coded the CDF in a library I’ve already got installed. It’s not knowledge but work that’s the limiting factor. :) Presumably this limits your prior quite a lot already, there being only so many commonly used math libraries.

RolfAndreassen 31 Jul 2009 16:39 UTC
1 point
in reply to: Eliezer Yudkowsky’s comment on: The Hero With A Thousand Chances
Out of curiosity, have you played the games, or did you just come across that particular quote and approve?

RolfAndreassen 2 Aug 2009 20:00 UTC
4 points
on: Pain
Very interesting question. I’ll have a go at it, although these are 30-second thoughts: Pain is a warning system, very strongly correlated with damage and/or danger of death. Damage is bad, hence when we feel pain we want to stop whatever is causing damage; this has the incidental side effect (usually) of stopping the pain as well, so it feels as though we want to get rid of the pain—in other words, it feels as though the pain is bad, although what we really want is to stop the damage. A fitness-maximiser would want to stop the damage without the intervening step of pain.

With that said, though, I appear to have moved the question without really resolving it, because why is damage bad? Why is death bad? I don’t think these can be answered without appeal to the plain utility functions: DO NOT WANT. So at best I’ve resolved the problem with masochism (edit to add: as a counterexample to people wanting to avoid pain), in that it’s a rare BDSM scene which intentionally inflicts actual damage—piercings, at most. Accidents do happen, but intentional damage is very rare.

RolfAndreassen 9 Aug 2009 22:50 UTC
2 points
on: Dreams with Damaged Priors

But with a damaged intuition for the prior probability, my mind didn’t successfully >check itself, or even notice a problem—didn’t get as far as asking “But what is the >prior probability?”

Ok, but in this case it seems that you took the correct action—don’t worry about it—and so it is not clear if this is a valid criticism. Even in terms of application of scarce brainpower, it seems to me that dismissing further worry by means of “low prior probability” is about the same amount of work as dismissing by the likelihood ratio, so once your dreaming mind has started considering the problem, there is no course more optimal than dismissing it and the path taken is not relevant. No?

RolfAndreassen 27 Aug 2009 17:01 UTC
2 points
on: Mathematical simplicity bias and exponential functions
Comparing a made-up exponential to a process that no scientist who knew anything about radioactivity would expect to model with anything but a sum of coupled exponentials is a bit of a straw man. There’s a bias to simplicity, certainly, but there’s not that much bias!