If it's worth saying, but not worth its own post, even in Discussion, it goes here.

New Comment
206 comments, sorted by Click to highlight new comments since:
Some comments are truncated due to high volume. (⌘F to expand all)Change truncation settings

Just a meta-comment for admins. The "Sequence Reruns" tag in the discussion section is now so common relative to the other tags, it's forced all the others in the tag cloud in the sidebar to the same relative size. That seems to be defeating the point a bit.

The tagging system in general is underused which is a shame,

While I didn't think much of the discussion in the recent creepy thread, I'm very much enjoying a series on a related subject written by Yvain.


I strongly recommend the whole blog - that guy should post on LessWrong or something!

Those posts are followed by:


I really liked the series. He should make a discussion post on LW linking to these with some commentary. I he doesn't I think I will. What he shouldn't do is make a neutered special needs padded "safe for LessWrong" version.

I think it would be better if some of the emotional appeals and personal elements were removed. ---------------------------------------- Maybe he should make a steroid-injected high-tier cutting-edge "too controversial for LiveJournal" version.
If it's safe enough for LiveJournal, I expect it to be safe enough for Less Wrong. He does a nice Eliezerish job of slowly easing people into ideas that they otherwise might not agree with.
Eighth is up.
Nine is here.
I don't think the example with the beggars in India provided the same insight to me that it did to Yvain. Mainly because tourists -- consistently, articulably -- don't want anyone to ask them for money, while women (I think?) do want (some) men to approach them with romantic overtures, and ostensibly filter them by certain characteristics. Tourist: don't ever ask me for money. So don't do things that make it harder for me to turn you down for money and then ask for money. Women[1]: don't ask me for romantic interaction unless I like you (or will like you), and if you get a false positive you're a f\**ing terrorist who did a thousand things wrong that I would have found charming if I liked you*. In other words, once you accept that some approaches of a certain type are desired, you have to accept that some will make that type of approach without it being desired, and therefore not regard such instances as atrocities. The tourist doesn't have this problem: s/he doesn't want begging at all. If there were a tourist that actually wanted to be panhandled but only from "awesome enough" people, and gave such people money I would expect that they'd have the insight to recognize that, "Well, some people will panhandle me without being awesome enough, and that doesn't mean they violated any rules. Here's a clear list of things that I regard as awesome, and here's what I do to mean that you should really stop revealing awesomeness and go away." [1] of the type who are most vocal in the creepiness threads Edit: I spoke too soon. Yvain addresses the above in the second meditation, but in the (more insightful) comparison to how "some tourists want their fortunes told". And indeed, some people want to be telemarketed to, and some want to be spammed. So what exactly makes some advances wanted and others not? I tried to address that issue here a while back, but all I got was resentment at the comparison to salesmen and telemarketers (oh, and a "you don't have the right to respo
I found them very interesting, though some of his statements and implications are naive or disingenuous. (As for Hanlon's razor, I assign a higher prior for the former and I'm not sure which way the evidence points overall. EDIT: after reading the ninth post of the series, it was definitely the former -- or that guy deserves an Academy Award.)
Awesome. Yvain should post these here. Or if he doesn't, he should at least appear here so we can upvote him for these...

Today I learned gwern.net has passed 1 million page-views. :)


Activist and Less Wrong user Aaron Swartz has been charged with 13 felonies for downloading millions of academic articles from JSTOR.

I'm trying to find the existing research on the topic Paul Graham discusses in this article (regarding the relative merits of programming languages in footnote 3 and surrounding text) and which EY touches on here (regarding Turing tarpits).

Basically, within the realm of Turing-complete languages, there is significant difference in how easy it is to write a program that implements specific functionality. That is, if you want to write a program that takes a bunch of integers and returns the sum of their squares, then it's possible to do it by writing in machine code, assembly, brainfsck, BASIC, C, Java, Python, Lisp, but it's much easier (and more concise, intuitive, etc) in some than others.

What's more, Graham speculates there's a ranking of the languages in which programmers too comfortable in one of the languages "don't get" the usefulness of the features in languages above it. So BASIC-addicts might not appreciate what they can do with recursion or structured programming (i.e. by abolishing go-to statements); C-addicts might not appreciate what they can do with functions as first class objects, and Python-addicts might not appreciate what they can do with Lisp macros.... (read more)

Graham speculates there's a ranking of the languages in which programmers too comfortable in one of the languages "don't get" the usefulness of the features in languages above it. So BASIC-addicts might not appreciate what they can do with recursion or structured programming (i.e. by abolishing go-to statements); C-addicts might not appreciate what they can do with functions as first class objects, and Python-addicts might not appreciate what they can do with Lisp macros.

If "enlightening people about better programming languages" ever becomes a higher priority than "enlightening people about superior status of X language users", I think a good strategy would be to explain those possible insights in a simplest possible form, without asking people to learn the guru's favorite programming language first.

For example, to show people the benefits of the recursion, I would have to find a nice example where recursion is the best way to solve the given problem; but also the problem should not feel like an artificial problem created merely for the sake of demonstrating usefulness of recursion.

I can use recursion to calculate the factorial of N... but I can us... (read more)

Have you dug around the publications of Alan Kay's Viewpoints Research Institute? They're trying to really push the envelope with the expressive power of highly customized programming languages. ETA: A LtU discussion thread about a recent VPRI progress report.
I haven't heard of any studies in that direction, although a few people try to do find something, like "how long are the programs comparatively" etc., similar to this quickly googled IEEE paper. I assume that because * programming languages are used by humans, and * most of a programming language's quality is based upon its actual effects on humans in the target group, and * for real work we value long-term effects such a study is unfeasible, i.e. too much effort, too expensive. And probably not that much related to programming language features (as most popular languages converge to some degree, especially on the most important axises). Also, "fewer symbols/lexemes/special-purpose-constructs" is an advantage only under certain conditions, meaning: the question asked may very well already determine the answer.
That's the short version. The full paper is here. I found it while looking for a similar comparison that I remembered seeing mentioned several times when I had been interested in Common Lisp and it turned out to be a follow-up to that. Oh, and those things actually looked at time spent programming, so they didn't measure only silly things like program length.
Why is program length a silly thing?
It was excessive to call it "silly" but program length still seems very imperfect way to measure the ease of writing programs in a given language. Better to directly measure things like programming time or number of bugs.
It's silly when you're measuring it in "lines of code", because "line" is a somewhat arbitrary construct, for which "chunks of text delimited by newlines" is a worse approximation than most people think. (Quick proof: in many languages, stripping out all the newlines yields an equivalent program, so that all programs are effectively one-liners.)
Then it's a good thing I didn't measure it that way, or use that term in this entire thread! Whenever I did refer to measures of program length, it was with constructions such as:
Thanks for digging that up! Also, I did not want to imply that only silly things are measured, but rather that the most interesting questions are still unanswered due to various constraints.
Do you mean only programming languages or programming languages plus the commonly available standard libraries? Some programming languages are very simple and powerful (LISP, Haskell), and some provide a large standard library and other tools to make it easy and straightforward to get things done in specific problem spaces (MATLAB, VB). The most powerful and concise language I can imagine would be in the language of set theory with a magic automated solver: Define the domain of a problem space and a predicate for solutions and the result is the set of solutions. A standard library built on such a language would consist mostly of useful definitions for converting real world problems into mathematics and back again. I think most programming languages try to approximate this to some degree, but the programmer is necessary to fill in the algorithmic steps between the problem space and the solution space. The fewer explicit steps the programmer has to specify to write a function, the more concise and powerful the language. The fewer definitions and functions necessary to convert real world input into a mathematical model and back to real world output, the more concise and powerful the standard library is.
Graham addresses the point you're making about the difference between the language being powerful vs it having a large standard library. As he says in the link, by his metric, two languages are "at the same level" if they only differ by one of them lacking a library function that the other has. His test for proving that language A is "above" B is the question: "To do a task in B, do you have to write an interpreter for A (or language on its level)?" For example, adding recursion to Basic requires going beyond writing one more library function.
After thinking about this for (a long) while, I think the most powerful languages would then be the ones that are internally extensible and allow modification of their own syntax and semantics. For instance BASIC would become a far more powerful language than C or LISP if it had a function for altering the interpreter itself. Certain versions of BASIC effectively had this by allowing functions to be written in machine language and then executed. Theoretically, the machine language snippets could extend the interpreter to add structured programming features or first class functions. The same could potentially be done with C by just including the Tiny C Compiler in the source (as both the compiled functions and C strings containing the source) and then reflexively recompiling the compiler to add features. What is most interesting to me is that making a language that can modify itself requires a complete definition of its execution environment in order to be safely extensible. The most powerful languages would have to fully and formally define their semantics as well as their syntax and make both accessible within the language so that extension of the syntax could could safely extend the semantics to match. In other words BASIC with assembly language is not enough if you don't know everything about the hardware and the interpreter you start with. From my CS student days I seem to recall reading that few programming languages have rigorously defined semantics ("This feature is implementation specific" and "This behavior is undefined" appears far too often in most specifications). The closest thing I can find is the Gödel machine, but that's defined directly in terms of a Turing machine, afaict.
I don't have any answers, but I'm also interested.

Learning to lose when you aren't Harry Potter-- an American studies ping pong in China.

The new Richard Carrier book, Proving History, is fantastic. Basically it's an introduction to Bayesian thinking for people who think in words. You'll enjoy it.

I think this is worth it's own post but in light of my last discussion catching fire and burning to the ground, I have decided to request a critique on this one before posting in discussions:

Cryonics Moral Dilemma

Since joining LessWrong, I've been thinking about cryo a lot, and have encountered a dilemma:

According to GiveWell, "We estimate that giving a few thousand dollars to AMF likely saves a person's life." (They do malaria bed nets if you're not familiar).

Cryo costs tens of thousands of dollars, and it's not guaranteed to save even one life.

I don't see how I would ever justify signing up, myself, unless I show that I'm capable of making a large enough difference in the world that rescuing my difference making abilities justifies the risk and cost.

This also means "Reddit, help me find some peace I'm dying young" is a cute puppy dog cause. :/

Does anyone relate? What are your thoughts?

please critique the proposed discussion post

This idea has been covered on Less Wrong before. I'll spend the next minute looking up some links. EDIT: Years saved: Cryonics vs VillageReach Against Cryonics & For Cost-Effective Charity There's already discussion about cryonics and charitable giving in the Reddit, help me find some peace I'm dying young thread. There is a discussion thread in Normal Cryonics about charity vs. cryonics. See in particular this comment.
Wow okay. I didn't expect to find such good arguments. I am still not adjusted to the intelligence level here. Well, different new discussion idea then.
For me the argument is the same as for why I don't live a pauper and give all my wealth to charity: Selfishness. I know that I can feed someone for a long time on the money I spend on a trip somewhere, and I still prefer to take the trip. I will spend a lot more money on myself than I will on a friend, and a lot more on a friend than on a stranger. Note: I am not signed up for Cryonics.
If cryonics works, then money spent on cryonics is much more of an investment than money spent on conventional charity. Several million people die every month. Malaria nets can only stop a small fraction of that, no matter how many are made, but cryonics can stop almost all of it - if it works. Anything done in support of cryonics in its fledgling form will help it to scale up. The future won't revive you because it needs you to solve the Y3K problem, but we also don't save children from dying in order that they can go back to work the next day. Cryonics is a way to stop a life from being cut off, with the side effect that the cryonaut wakes up as a mere human in a transhuman world. If it's a friendly place, they'll have a chance to grow into their new world as an equal and a participant.
Grats, you are catching on :)

(: Thanks, Shminux.

I have finally gotten the ass-kicking I needed. Though not especially in my elitism thread, it was spread out... Wedrifid showed me arguments good enough to corner mine. Kindly provided a wonderfully devastating critique of my poll. Gwern's website shows that he's so well-read that I felt like an idiot. Eliezer's "The Magnitude of His Own Folly" depicted a deep acknowledgement of the terrible nature of reality that I found moving because it made him neither paranoid or unambitious - I relate to this but I haven't seen anyone like that before. You always seem to be there to say something snide, making my overconfidence think twice while Morendil typed me up a refreshing batch of sanity.

These are exciting.

I haven't felt so much respect and faith in humanity for a long time.

I was getting apathetic because of it.

Now my self-confidence is right about where it should be.

I decided to commit to reading the major sequences, and I'm considering reading them all. I previously did lots of things like learning about logical fallacies and razing my cached thoughts years ago, so these aren't as dense in new information as they'd be otherwise, but I'm learning to communicate with you guys and I'm enjoying Eliezer's brilliance.



Why didn't someone tell me about this earlier?

Geoff Anders just showed me this PowerPoint prepared by U.S. Air Force's Center for Strategy and Technology, the same group that produced this bombastic 'future of the air force' video.

Slide 18 makes a point I often make when introducing people to the topic: the military's policy assumption is that humans will always be in the loop, but in reality there will be constant pressure to pull humans out of the loop (see e.g. Arkin's military-funded Governing Lethal Behavior in Autonomous Robots). The slide concludes: "In fact, exponential technological chan... (read more)


Almost a Century Ahead of The New York Times

Many of the influential thinkers, prestigious publications, and important articles of that bygone era are almost totally unknown today, even to many specialists, and the vacuum produced by that loss of historical knowledge has often been filled with the implied histories of modern Hollywood movies and television shows, some of which are occasionally not totally accurate or realistic. Indeed, a casual perusal of the major writings of the past often seems somewhat akin to entering a science fictional alternate-re

... (read more)
So, Cochran and Harpending have been posting to their blog about genetic noise and parental age. Given modern data, this is actually a rather important result with a lot of wide-ranging implications. Turns out, people thought it was significant 50 years ago, and it's mostly lain dormant since then.
In genetics, it seems to be the rule that speculation far outpaces what can actually be known. Countless times I read these papers and they go 'as speculated by X 50 years ago...' (where X is usually Darwin or Fisher). I understand there's some question as to even whether Mendel's peas showed the laws he wanted them to show! Which would indeed exemplify the theory outpacing the practice.

Evidence is building that High intensity interval training, e.g. Tabata sprints, is more effective at physical conditioning than low intensity endurance techniques. In terms of weightlifting, "low-rep, high-weight" workouts seem to be better than "high rep, low-weight" workouts.*

I wonder if something analogous is true for mental training. E.g., will you improve mathematical ability faster by grinding through a bunch of relatively easy problems, or by spending a shorter amount of time mentally exhausting yourself on problems that push yo... (read more)

How about 'deliberate practice'? I'm fairly sure that it implies that you're working on a problem that challenges you and pushes your limits.
I remember an unconference now from the July minicamp on deliberate practice now. IIRC the speaker suggested something similar to beoShaffer's comment.
I forgot to mention that I was basing my info on a keynote speaker that I suspect may have done the minicamp unconference. Was the speaker a female psychology professor who made numerous movie references.
Nope, it was a male. I think it was Mark E, known around LW as Mark E (I don't recall how to say/spell his last name, I just remember it being somewhat complicated and that it's also part of his LW name)
Nevermind then.
From what I understand deliberate practice would generally favor the small number of hard problems, especially for building overall mathematical competence/your ability to tackle hard problems. However, doing the easy problems in a challenging way, like trying to do them as fast as possible while still maintaining a high standard of accuracy, would also lead to improvement, particularly for you ability to do that specific type of problem quickly and accurately.
What's the basis behind HIIT? If I remember correctly, it's that the high intensity activity kicks your metabolism up a notch, continuing to burn calories / seem active for a significant period after the training is officially complete. Is there a similar mechanism for learning and memory? There's solid evidence that spaced repetition- like in the Saxon method- is demonstrably better than doing something once and moving on with little review. In general, it seems like practice is a very important part of mathematics ability. There are also time-based effects for learning things before going to sleep- but I'm not sure how practical using those would be.
Off-Topic Nit-Picking: "physical conditioning" is a very general term. For instance: Is evidence building that Tabata sprints are more effective for preparing for a 100k ultra-marathon? Of course competitive runners do some sort of interval training, and -- if information on The Internet (reddit) is to be believed -- runners do not train the full distance. And if basic health and looks is your goal, running is probably not the most time-efficient (of even effective) way for doing it. But this "endurance is all wrong" meme is overshooting it a bit...

Good news for EY: Akinator can recognize him and unequivocally believes that he's 'famous'. Bad news for EY: Akinator also believes that he uses 'green energy' to 'power up'. I'm not sure if that refers to an environmentally clean power source or literally green supervilllain rays of death, like those of the Necrons in WH40k. Probably the latter. In any case, it hardly improves SIAI's public image :(

I'm responsible for Akinator knowing who he is, having a picture of him, and in part for believing he's famous. I don't know where the energy's from. Edit: Akinator still asks about green energy but now expects a "no".
I tried and didn't get Eliezer. This is possibly because Akinator doesn't believe that any actual famous people are personally known to the people playing the game? I guess it would have had an even harder time if I'd played back when I would've had to say "yes" to "does your character live with you?". Edit: Told it it was wrong and hit continue; still got it wrong. When I told it it was wrong again it wanted me to pick off a list, and admittedly Eliezer was on the list!
I got Eliezer (the page's second guess) despite saying "no" to the green energy question.
If you click on "Game Report" at the end, it will tell you which answer it expected for each question, given who the person is. (It is capable of guessing the correct person even with a few "unexpected" answers.)
A year ago, it specifically got Harry James Potter Evans Verres from Methods of Rationality.
I tried for Luke and he wasn't a result. After the last question, Eliezer was its second guess. (Edit: this is not directed at Multihead in particular.)

Kind of a stupid question:

It's a truism in the efficient charity community that when giving to charity, we should find the most efficient group and give it our entire charity budget; the common practice of spreading donations among groups is suboptimal. However, in investing it's considered a good idea to diversify. But it seems that giving to charity and investing are essentially the same activity: we are trying to get the highest return possible, the only difference is who gets it. So why is diversification a good idea for one and not the other?

It's a truism in the efficient charity community that when giving to charity, we should find the most efficient group and give it our entire charity budget; the common practice of spreading donations among groups is suboptimal. However, in investing it's considered a good idea to diversify. But it seems that giving to charity and investing are essentially the same activity: we are trying to get the highest return possible, the only difference is who gets it. So why is diversification a good idea for one and not the other?

If you are attempting to maximise expected returns from your personal investment you would not diversify (except within resources that have identical expected returns). However with personal investments you have some degree of risk aversion. That is, you don't value money linearly all the way from 0 to $10,000,000 and so splitting the investment between multiple stocks gives higher expected utility even though the expected returns in $ will be slightly lower.

This differs when it comes to charitable giving because it is assumed that your personal donations aren't sufficient to change the marginal utility significantly. Personally owning $10,000 rather than $0 is much more useful than owning $20,000 instead of $10,000 but after you give $10,000 to The Society For Cute Puppies And Mosquito Nets the value of giving another $10,000 to TSFCPaMN has probably barely changed at all. Diversifying becomes important again when you have enough financial power to change the margin all on your own.

You've pinpointed it: the only difference is who gets it. When investing, diversification as the receiver of the return is useful because you'd rather gain slightly less than often lose everything. When ... living, diversification as the receiver of the return is useful for the same reason.

When investing, you'd like your buyers to diversify... but there's only one buyer, so that buyer needs to diversify. But when giving charitably, the world would like its buyers to diversify, and there are lots of buyers. Assuming its buyers are sufficiently independent, the world gets enough diversification just because its buyers make different decisions. So as long as sufficiently many people make different charitable giving decisions than you, feel free to buy only what you think are the most efficient charities.

The world doesn't care how much you help it, the world only cares how much it gets helped overall.

I'm not sure if the arguments for diversification in investments actually apply to charity. You want to diversify your investments because you're risk averse. I would not, for example, bet $1000 on a coin flip; losing $1000 is more painful to me than gaining $1000 is pleasurable. In other words, your utility is not linear in money in your bank account. However, for charity, I think it makes perfect sense to have linear utility in money donated to charity. If you value saving two lives twice as much as saving one, and the cost per life saved is constant, then you should value each dollar given to charity as much as the last. Given that, you shouldn't really care about variance; you can focus on expected returns. As such, I don't think you should diversify charity donations at any scale, personal or on a worldwide scale; just donate to the most efficient charity, and then when that charity becomes less efficient, donate to the newest most efficient.
Now that I have read your answer, it seems obvious in retrospect. Very nice, thanks!
The difference is very simple. Is it better to have $100,00, or a 30% chance of $1,000,000 and a 70% chance of being homeless. Obviously the former Is it better to save 1 life, or have a 30% chance of saving 10 lives and 70% chance of doing nothing. Obviously the latter.
It seems true that when investing, you're trying to get the highest return possible, in terms of a single value measured in currency. I've never understood why it should also necessarily be true with charity. It seems often to be an unexamined assumption, and may be reinforced by using terminology like "utilons" that appears to be begging the question. Someone who donates both to the mosquito nets effort in Africa, and to the society which helps stray dogs and cats in Michigan, is not necessarily being irrational. They just may be perceiving the two benefits to lie on incomparable axes. They may be caring about helping Africans and helping stray dogs simultaneously, in different ways that are not exchangable to each other. The familiar objection is: "Sure they are exchangable; everything is exchangable into utilons; if you don't see a clear rate of exchange for your own preferences, that just means you still ought to estimate one given your imperfect knowledge, and act on it". But I don't see why that should be true. Certainly most of our spending is done on axes that are incomparable to one another. We have needs along those axes that we do not normally consolidate to one "most efficient" axis, even after the minimal requirements are met. Investing is the activity that's the odd one out, here - and one of the reasons it is is precisely that we don't care much which of the companies we invest in brings us profit. It seems odd that charity should so unequivocally stand along with investing as an exception. If charitable giving is not an exceptional way for us to spend money, the idea of a single currency becomes difficult to support, because if charity must be so streamlined, why not all other activity? In other words, sure, you can criticize someone for helping stray dogs by saying their money could be saving lives in Africa instead; but is that very different from criticizing them for buying a large color TV, when their money could be saving lives in Africa ins
Charity falls in the same category as investing to the extent that you care about the effectiveness of the different charities (as opposed to feeling good about yourself, for example). Here's why. For the sake of simplicity, suppose that you have $2000 to give to charity, and $1000 can either save a child in Africa or a dog in Michigan. For now, we assume that you care about the number of children and dogs saved. If the charities currently have enough money to save 999 dogs and 999 children, then preferring an even split to a $2000/$0 split means preferring 1000 dogs and 1000 children saved, to 1001 dogs and 999 children. Nothing wrong with this, by itself. However, we aren't precisely certain about these numbers; and if the charities have enough money to save 1000 dogs and 998 children, then preferring an even split to a $0/$2000 split reveals exactly the opposite preference. This is a problem. In general, as long as our uncertainty about how much the charities are doing is much greater than the impact of our own donations, a similar thing happens. It's easy to have enough information to prefer a $2000/$0 split or a $0/$2000 split above all: for instance, if you think it's best to have 1000 children and 1000 dogs saved, and currently there's money to save about 500+/-100 children and 1500+/-100 dogs, you should definitely donate all your money to the children charity. But having enough information to definitely prefer an even split is nearly impossible. The best we can do is consider the case where the difference is probably small, so you shouldn't really care one way or the other. This is probably rare, though, and even that doesn't argue in favor of an even split. Now, obviously our assumption that we only care about the totals is unrealistic. But then, the usual argument runs, maybe we should figure out how much we care about the totals, and efficiently distribute that portion of our money (most likely, only to one charity). After that, the remaining money
If you decide that donating 1$ to mosquito nets and 1$ to stray dogs is better than 2$ to one or 2$ to the other, then you have in fact performed a comparison between those three actions. If the type of good generated by mosquito nets is one axis and the type of good generated by saving stray dogs is another, then the scalar-valuedness of utility isn't about the axes, it's about comparing any given point in that 2-D space with any other point. The alternative to being able to compare things isn't some decision process other than comparison. The alternative is to not have preferences about the state of the world at all; to say that there is no such thing as a "right thing to do" in a given circumstance. Expected utility does apply to all activity.
Granted, preferring one particular 2D point to another may be read as running a scalar-valued comparison function on the 2D space (such a reading is not without problems, e.g. because real people's preferences may not be transitive, but let's ignore those details). However, from the existence of such a function it does not follow that "we should find the most efficient group and give it our entire charity budget" - this being the claim the universality of which I was contesting.
Agreed. To derive that you would also need a smoothness constraint on said function, so that it can be locally approximated as linear; and you need to be donating only a small fraction of the charity's total budget, so as to stay within the domain of said local approximation. I assert that the smoothness property is true of sane humans' altruistic preferences, but that's not something you can derive a priori, and a sufficiently perverse preference could disagree.
You're solving essentially a global optimization problem; what use is (the existence of) a local linear approximation? If the utility function happens to be the eminently smooth f(x,y)=xy, then under the constraint of x+y=const the optimal solution is going to be an even split. It's possible to argue that this particular utility function is perverse and unnatural, but smoothness isn't one of its problems. You don't even need contrived examples to show that utility functions do not admit their maxima along one axis. My other point was that charity may not be easily distinguishable from other types of spending[1], and our normal utility functions definitely don't have that behavior. We do not, among different types of things we require/enjoy, pick out the "most efficient" one and maximize it alone. [1] As another example of that thesis, consider the sequence: I buy myself a T-shirt - I buy my child a T-shirt - I pool funds with other parents to buy T-shirts for kids in my child's kindergarden, including for those whose parents are too poor to afford it - I donate to the similar effort in a neighbouring kindergarden - I donate to charity buying T-shirts for African kids.
Yeah, but unless you actually end up at that point, that's hardly relevant. If people donated rationally, we would always be at that point, but people don't, and we aren't. We're normally only dealing with one person. If you play videogames, you quickly get to the point where you don't want to play anymore nearly as much, so you do something else. If you save someone's life, there's still another guy that needs saving, and another guy after that, etc. You can donate enough that the charity becomes less efficient, but you have to be rich and the charity has to be small. Also, consider: If you wanted a shirt, and I bought you one, you'd stop wanting a shirt and spend your money on something else, just like if you bought the shirt. If you wanted to donate $100 to X charity, and I told you that I already did, would you respond the same?
I don't understand how what you just said relates to my example. To recap, I meant my example, where the maximum is at the even split, to refute the claim that any smooth utility function will obtain its maximum along one "most efficient" axis. The whole argument is only about the rational behavior. While this is true, and does point at an interesting difference between charity and many other behaviors, it can't isolate charity, not by a long shot. There are many, many things we do that we stop doing not because of satiety or exhaustion, but because of other priorities. To give the first example that comes to mind, a personal one, I'm learning piano and I also play table tennis. I enjoy both activities immensely and would like to do either of them a lot more (but can't because of other commitments). There's no question of satiety of exhaustion at the level I currently invest in either. I could stop doing one of them and use that time for the other, but I explicitly don't want do to that and consider that an inferior outcome. I don't think this preference of mine is either irrational or very unusual. Here's a closer "personal spending" analogy to charity: I commit to putting aside $500 every month for a future downpayment on a house (a goal far in the future). A family friend gives me an unexpected present of $500, putting it right into the fund. Am I likely to forego my own deduction this month and use it for other fun things? Depends on the kind of person I am, but probably not.
Kindly's comment gets it right. It's not about satiety. If you're a consequentialist and care about the total amounts of money donated to each charity, rather than about how much you donated, then the decision in favor of the equal split must be very sensitive to the donations of others like you. That's the relevant difference between selfish spending and charity.
You only control a tiny portion of the money that gets donated to charity. If there's currently an equal amount of money donated to each charity, the ideal action would be to donate equally to each. If the difference between the amounts exceeds the amount you donated, which is more likely the case, you donate to the one that there's been less donated to. For example, if one charity has one million dollars in donations and the other has two million, and you donate a hundred thousand over your life, you should donate all of it to the charity that has a million. I doubt that. You might still have fun doing each more, but not as much. If you chose to learn the piano before, but now choose to play tennis, something must have changed. If nothing changed, yet you make a different decision, you're acting irrationally. Why is that analogy closer? It looks like it's in far mode instead of near mode, and the result is more controlled by what's pretty than what makes you happy. For example, if you got a $500 a month raise, you likely wouldn't save it all for the downpayment, even though there's no reason to treat it differently. If you got a $500 a month pay cut, you almost certainly wouldn't stop saving.
I've made a post with an analysis of this situation.

Reposting a comment I made on Yvain's livejournal:

There's a standard argument about "efficient charity" that says you should concentrate all your donations on one charity, because presumably you have preferences over the total amounts of money donated to each charity (not just your own donations), so choosing something like a 50/50 split would be too sensitive to other people's donations.

I just realized that the argument applies in equal force to politics. If you're not using "beliefs as attire" but actually care about politics, your participation in politics should be 100% extremist. That's troubling.

You might be an extreme centrist. Or an extreme pragmatic. Not all extremists are the "take some idea to its (il)logical conclusion and start blowing things up" type.
I believe the point is that while your personal beliefs may lie at any point in some high-dimensional space, if you're getting involved in politics in some anonymous way you should throw all your support behind the single "best" group, even if, like in two-party politics, that means supporting a group you have significant differences with. Non-anonymity (nonymity) changes things, leading to behavior like lobbying multiple parties. I don't really find it that disturbing, but it does get a little weird when you remember how bad humans are at separating acts from mental states.
Impact of charitable donations is, at least within the domains that most people can give, directly proportional to the size of the donations. It's not at all clear, however, that extremist participation in politics produces a greater impact in the desired direction than casual participation. I think that in some cases, it probably does, whereas in others it does not.
It probably depends of the decision process you're trying to influence: * If you're voting for a candidate, you don't have any incentive to vote in a way more extreme than your preferences - with more than two candidates, you can have strategic voting which is often the opposite incentive, i.e. voting for a candidate you like less that has more chances of making it. * If a bureaucrat is trying to maximize utility by examining people's stated preferences, then you can have an incentive to claim extreme preferences for the reasons Yvain gives. Informal discussions of what social norms should be look more like the second case. Elected politicians have to deal with the two systems, on one side they want to take a moderate position to get the maximum number of voters (median voter etc.), on the other hand once elected they have an incentive to claim to be more extreme when negotiating in their constituents' interest.
Could you spell out what you mean by extremist, and how the analogous argument goes? If there are three candidates, then yes, you should give all your support to one candidate, even if you hate one and don't distinguish between the other two. But that hardly makes you an extremist. I don't see any reason that this kind of argument says you should support the same party in every election, or for every seat in a particular election, or that you should support that party's position on every issue. Even if you are an extremist and, say, want to pull the country leftward on all issues, it's not obvious whether equal amounts of support (say, money) to a small far-left party will be more effective than to a center-left party. Similarly, if your participation in politics is conversation with people, it's not obvious that always arguing left-wing positions is the most effective way to draw people to the left. It may be that demonstrating a willingness to compromise and consider details may make you more convincing. In fact, I do think the answer is that the main power individuals have in arguing about politics is to shift the Overton window; but I think that is a completely different reason than the charity argument. ---------------------------------------- and then I looked up your comment on LJ and the comment it replies to and I strongly disagree with your comment. This has nothing to do with the charity argument. Whether this argument is correct is different matter. I think the Overton window is a different phenomenon. I think the argument to take extreme positions to negotiate compromises better applies to politicians than to ordinary people. But their actions are not marginal and so this is clearly different from the charity argument.
I agree with everything in your comment. "Extremist" was a bad choice of word, maybe "single-minded" would be better. What I meant was, for example, if success at convincing people on any given political issue is linearly proportional to effort, you should spend all your effort arguing just one issue. More generally, if we look at all the causes in the world where the resulting utility to you depends on aggregated actions of many people and doesn't include a term for your personal contribution, the argument says you should support only one such cause.
2Wei Dai
But this isn't at all likely. For one thing you probably have a limited number of family and friends who highly trust your opinions, so your effectiveness (i.e., derivative of success) at convincing people on any given political issue will start out high and quickly take a dive as you spend more time on that issue.
I'm inclined to agree. A variant of the strategy would be to spend a lot of time arguing for other positions that are carefully selected to agree with and expand eloquently on the predicted opinions of the persuasion targets.
Yes, that is the charity argument. Yes, you should not give money both to a local candidate and to a national candidate simultaneously. But the political environment changes so much from election to election, it is not clear you should give money to the same candidate or the same single-issue group every cycle. Moreover, the personal environment changes much more rapidly, and I do not agree with the hypothesis that success at convincing people depends linearly with effort. In particular, changing the subject to the more important issue is rarely worth the opportunity cost and may well have the wrong effect on opinion. If effort toward the less important issue is going to wear out your ability to exert effort for the more important issue an hour from now, then effort may be somewhat fungible. But effort is nowhere near as fungible as money, the topic of the charity argument.
Value of information about which political side is more marginally valuable makes unbiased discussion a cause that's potentially more valuable than advocacy for any of the political sides, and charities are similarly on the same scene. So the rule is not "focus on a single element out of each class of activities", the choice isn't limited to any given class of activities. Applied to politics, the rule can be stated only as, "If advocacy of political positions is the most marginally valuable thing you can do, focus on a single side."
Yeah, I agree. I wonder how many people would subscribe to the rule in full generality.

How many threads and discussions have we had about LessWrong readers joining or participating in online classes? How many about forming study and reading groups for such classes or textbooks? Please help me complete this list. I'm interested in getting a quick idea of what works and what doesn't in trying to educate parts of the community as whole. I think it is relevant to the problem of our subculture not updating as well as some other efforts I'm currently working on.

Online Classes

... (read more)
I'm surprised no-one's thrown together an online course aggregator yet.
You mean something like Class Central?
Yes. It's a shame nothing like that already exists.

Can anyone recommend any books on signalling or on feminism that might appeal to a LWer?

I too am interested in books on feminism that might appeal to a LWer. I tried researching this, but I found one easy failure mode is to find books based on the assumption that everyone knows what feminism is about and agrees 100%, and what the reader wants is a list of people who were feminists at various times in history. The angle I was more interested in was "feminists believe X because Y", or "you, foolish person who is not a feminist, should be a feminist because Z".
Also interested in books that might appeal to a LWer. This Lukeprog post and comments from a while ago might be of interest.
On signalling: * Comeuppance: Costly Signaling, Altruistic Punishment, and Other Biological Components of Fiction by William Flesch attempts to explain what makes certain plots, or fiction at all, enjoyable through a signalling lens. * The Mating Mind and Spent by Geoffrey Miller delve into how signalling concerns pushed us evolutionarily into who we are and then how our evolved tendencies are disconnected with modern consumer culture. * Codes of the Underworld: How Criminals Communicate by Diego Gambetta
Your amazon links aren't working. Does Comeuppance give any clues about why Martin's Game of Thrones series is so wildly popular? It's a world with plenty of power maneuvering but little or no justice so far.
Thanks about the links. Flesch would argue it's largely because there is so little justice in the books. We're interested in tracking others through stories to see who deserves punishment. We remain emotionally involved to see what happens to them, and more injustices mean even more reason to keep tabs on what they're doing. Anticipation of justice is more satisfying than seeing justice itself.
I have to admit there's a character I want to see smashed flat. In normal fiction, he would be and I can't count on it happening with Martin, but it still feels like a reason to read book six, when and if it comes out. Does Flesch get into the difference between punishment and challenge? Either can be a strong hook for readers.

I've enrolled in 3 Coursera classes as a kind of warm-up for (possibly) going back to school to study computer science (if my start-up succeeds before then). They are:

  1. Introduction to Mathematical Thinking by Keith Devlin
  2. Learn to Program: The Fundamentals by Jennifer Campbell and Paul Gries
  3. Introduction to Logic by Michael Genesereth

Reply to this comment or PM me if you are interested in collaborating.

That's amusing. Usually I would say the value of the founder being present is much higher for a successful company than one that has failed. I would actually expect my freedom to pursue other avenues diminish as my success in my current avenue grows. Do you mean that your start-up, if successful, will pretty much run itself? Or that if it hasn't succeeded /yet/, then you will feel obligated to stay and keep working on it?
Schooling (for me) is as much consumption as investment. I'm merely saying that if my income significantly increases, then I will engage in more consumption (getting a degree in computer science, going on a pilgrimage in Europe, etc...). Is this really so strange?
A little. As your income increases, I expect your consumption to become more expensive in monetary terms, but as your business grows I expect the value of your time to increase and for your consumption patterns to become less expensive in terms of time. College is very expensive in terms of time. I'm not saying this is a bad choice, but it is one that surprises me. I'm still interested in the answers to my questions. Do you intend to sell your start-up, have it run itself, or abandon it? It seems like those options cover the gamut (I might consider requiring < 40 hours a week of your time to be "running itself"; if you're quite dedicated, you could probably fit being a full-time student in even with the start-up taking 40+ hours of your time, making that an alternative option).
Ah, I think I see the source of your confusion. If my start-up succeeds, then I plan to increase the time I spend doing it and schooling, since I currently work a full-time job and work on my start-up part-time. The relevant options are full-time work/part-time entrepreneur or part-time school/full-time entrepreneur.
Indeed, I misinterpreted you in multiple ways. My model went something like "Jayson_Virissimo is currently working 60-80 hours a week on his start-up. Once it exceeds ramen-profitability, he intends to scale back his efforts to become a full-time student." How very foolish of me!
No, but context and wording is sometimes everything. I assume that datadataeverywhere has imported some HN style startup discussion context here.
I'm thinking of doing all three, though I might skip 2 since I'm already doing things along those lines.
So far, 1 and 3 have lots of overlap. I'm not sure if that is good or bad yet.
I'm doing mathematical thinking as well. I have a two-person local study group; how large is the LW study group looking so far?
So far, it contains Curiouskid, khafra, and Jayson_Virissimo.
Should we set up a facebook group for this? How do you guys plan on communicating?
I've been taking their Quantum Mechanics and Quantum Computation course that is about to end. I would recommend it to anyone interested. The course doesn't do a good job emphasizing the linear algebra prereqs (basically the only prereqs of importance) and it sounded like a lot of people got frustrated with that early on. A college course in linear algebra definitely suffices and one could even learn this sufficiently for the course on your own if you were ready to dedicate some time to it.

Luck egalitarianism

Luck egalitarianism is a view about distributive justice espoused by a variety of egalitarian and other political philosophers. According to this view, justice demands that variations in how well off people are should be wholly attributable to the responsible choices people make and not to differences in their unchosen circumstances. This expresses the intuition that it is a bad thing for some people to be worse off than others through no fault of their own.

Luck egalitarians therefore distinguish between outcomes that are the result of

... (read more)
Well, we already know that (even in a deterministic world) there's a meaningful way to say that someone could have done something, despite being made of mindless parts obeying only the laws of physics. I think the notion of responsible choice is probably similar.

On LW we talk occasionally about "life-hacks" (simple but non-obvious ways of solving common problems/becoming more productive, etc.) However, these are often considered too off-topic for LW. I distinctly remember reading a long list of life-hack ideas on some website, and a lot of them seemed very promising, but I apparently never bookmarked it.

Is there any good place on the Net to find the more effective hacks? It seems like there are a number of easy-to-implement ideas out there that would help a lot of people, but they are not concentrated... (read more)

I've finished transcribing the classic sociology paper "The Iron Law Of Evaluation And Other Metallic Rules"; LWers or libertarians may enjoy it.

I posted earlier on the advantages of incorporating audiobooks into your study methods. One of the main problems I desribed was that there was poor selection with regards to audiobooks and particularly with regard to higher level subjects. I've recently found a way around this that makes using audiobooks even more of an obvious decision for me. I've started using text to speech conversion to make audiobooks from ebooks. The inspiration was from wedrifid.

Here is a sample of the best TTS voice I have been able to find. This method produces suprisingly high q... (read more)

I've been doing this since November last year and recommend it. My list of fully listened books has 109 entries now. I've found that an important thing in determining whether a book works well in text-to-speech form is how much of it you can miss and still understand what's going on, or, how dense it is. Genre-wise, narrative or journalistic nonfiction and memoirs make especially good listening; most popular nonfiction works decently; history and fiction are pretty hard; and scholarly and technical writing are pretty much impossible. A lot of writing on the internet, like blog posts, works well too. I have some scripts for scraping websites, converting them into an intermediate ebook form and then into text-to-speech audiobooks. If I encounter an interesting blog with large archives this is usually how I consume it nowadays. There's also obviously the issue of comprehension, which I'd say definitely is lower when listening than when reading. But, 1) literally at least 95% of the stuff that I've listened to I never would have read, it would either have sat on my to-read list forever* or I wouldn't have thought it was worth the time and effort in the first place 2) you can view this as a way to discover stuff that's worth deeper study, like skimming, 3) it takes less mental effort than reading, and 4) there are a lot of times when you couldn't read but can still listen, so it's not a tradeoff. There are also some interesting ways in which texts are more memorable when listening, because parts of the text get associated to the place you were in and/or the activity you were doing when you were listening to that part. Compared to traditional audiobooks, there's the disadvantage that fiction seems to be harder to make sense of in text-to-speech form, but other than that, you get all the benefits of traditional audiobooks plus it's faster** and you can listen to anything. * Whereas during the last year I've been getting to experience the new and pleasant phenomenon of
I don't have a much problem listening to to scholarly stuff what problems did you have with it? Most of the books I listen to recently probably count as scholarly. I don't think that my comprehension is lower than when I read conventionally, actually, I need to test it, though. What method do you use for converting blogs? Have you found a way of converting a whole blog at once or do you convert articles individually?
Well I had in mind how at one time or another I tried to listen to Inside Jokes, Folly of Fools, In Gods we Trust, Godel's Theorem: An Incomplete Guide to Its Use and Abuse, The Origin of Consciousness in the Breakdown of the Bicameral Mind and The Terrible, Horrible, No Good, Very Bad Truth About Morality and What To Do About It and quit each of them because it wasn't working, I was missing too much and wasn't enjoying it. Maybe "scholarly" isn't the best word I could have chosen to describe them, and maybe I was just doing it wrong, and should have just gone slower and concentrated better. The result of a converted blog is this. I just have to write a new parser for every new blog, which usually takes maybe 15 minutes, and the rest is automated.
Wow, those are some awesome books, thanks. I listen to the The Terrible, Horrible, No Good, Very Bad Truth About Morality and What To Do About It and at 3.8x speed and I think I understood it fine. I'd love to be able to convert blogs... Can't find any service to do it for me.

Random idea: LW would benefit from some good feminism articles. Someone with more money than me (CFAR) should incentivise creation of them with a suitable prize.

But on what? Question choice is hard. (My own suggestion would be something like hold a contest on finding and summarizing all existing criticism of Baumeister's Is There Anything Good About Men? and if there is no good criticism, come up with one's own; but that's just because I regard it as one of the more evidence-backed paradigms.)
We could start with an overview of different kinds of feminism.
this is totally a hard problem I agree.
Intrinsic motivation, pplz :D. Better than a prize might be paying someone to convince and help other people to want to write the article. Or simply doing that yourself, if you want to make it happen.
Primary takeaway: simple, visible theories are rarely completely correct, especially when they're formulated in response to primitive data. Commentary: "Species" is a wrong idea, and much of the start is uninteresting discussion of whether or not neanderthals were a "separate species." Taboo species, and everything becomes clear: they had a separate evolutionary history than the strain of homo sapiens that originated in Africa, but they could and did interbreed with that strain, and so current humans have genes from at least one of the African, Neanderthal, and Denisovan varieties.
I agree. I was shocked that a prominent anthroplogist would be so essentialist about the concept of 'species'.

The Spirit Catches Lia Lee, RIP

"First published in 1997, Anne Fadiman's book The Spirit Catches You and You Fall Down, a chronicle of a Hmong refugee family's interactions with the American medical system in the face of a child's devastating illness, has become highly recommended, if not required, reading for many medical students and health care professionals, over the past 15 years quietly changing how young doctors approach patients from different cultures. On August 31, with little publicity, Lia Lee, the young girl who inspired the book, af

... (read more)
1[comment deleted]

Movement toward taking a statistical approach to the quality of evidence from fingerprint similarity

Before you read that article, what was your opinion of fingerprint evidence? [pollid:74]

My first exposure to the idea that fingerprint evidence might not be all that good was in L. Neil Smith's The Probability Broach, in which the viewpoint character, a policeman, wonders whether all fingerprints really are unique, and also wonders whether the government might disappear people who had identical fingerprints.

(Surprisingly to me, the Salon article mentions tha... (read more)

Write-in: I believed it was among the more reliable forms of forensic evidence, but didn't believe the bombastic claims of absolute certainty.
The article doesn't actually contain any data saying that fingerprints are reliable. If I had to guess, I'd say that a (non-partial) fingerprint match had an odds ratio of around 10^7, a hefty 22 bits of info. Is there any data to contradict that? Or is this just the "but there's still a chance, right?" fallacy?
0[comment deleted]

Garbriel Kolko on the New Deal

If this is accurate, history is more complicated and less dramatic than usually thought, as is commonly the case.

The article is very limited as a history of the Great Depression - it says very little about why the economy got worse and better and does not include key words like "deflation", "gold standard", "Federal Reserve", or "monetary". One of the first things that Roosevelt did as President was to take the US off the gold standard (to put a stop to the deflation), and that was probably the most important thing that a country could do to deal with the Depression. See, for example, this graph, this section of a Wikipedia article, or this (much longer) Economic History Association article.
Still, if Hoover's policies weren't all that different from Roosevelt's, and most of Roosevelt's didn't make all that much difference, that's quite different from the usual account. Also, is it true, as claimed in the article, that WW2 got the US out of the Great Depression? This has never seemed plausible to me (get richer by doing a huge amount of non-productive work?), and wars don't usually seem to be great for the economy.
The article is overstating the similarities between Roosevelt's policies and Hoover's. The Wikipedia article which I linked covers it pretty well. Some of the main components of Roosevelt's policies were: * Ending the gold standard * Regulating the banking/finance industry, including creating the FDIC (which put an end to bank runs) and the SEC * Creating a safety net (e.g., Social Security and the precursor to food stamps) * Expanding & creating more public works / job creation programs (Hoover had some, FDR employed as much as 7% of the workforce) * Labor market regulations, including prohibiting child labor, creating a minimum wage, and establishing a 44-hour workweek You could divide the consequences of policies into three categories: getting the country out of the Depression (improving the economy, or making it worse), getting the country through the Depression (coping with the bad economy), and making lasting changes to the government and society. Most of these policies did more for the latter two. For getting out of the Depression, the most important things the government could do were 1) monetary policy, and 2) increasing total government spending and debt. Getting off the gold standard was the key step for monetary policy (there were also various missteps by the Federal Reserve, including many in the late 1920s and one in 1937). Hoover increased total government spending and debt somewhat and then stopped; FDR kept them relatively flat until WW2 broke out and then increased them massively.
Would it make a difference if the effects of the war was described as decreasing real wages (by freezing nominal wages in big inflation) and killing a bunch of men?
It's sort-of plausible, because the GD was due to a severe shortfall in overall nominal expenditure wrt. the prevailing level of prices and wages. The original cause of this shortfall was a series of mistakes in monetary policy; however, increased deficit spending in WW2 could have made up for it.

In the honor of our swanky new poll tech, help me determine which post I should write next. (By next, I mean "in parallel with a chapter-by-chapter review of Causality and my rationalist MLP fanfic," so no guarantees which gets done first.)

The main candidates are:

  1. Rereading An Intuitive Explanation of Bayes' Theorem, I was struck by how uninformative the introduction was. Why do you want to learn Bayes? Because it's cool! It seems like a post explaining what mindset / worldview would find Bayes useful might be a useful complement to that.

  2. About

... (read more)
Woah, woah, woah! Polls! And even with graphs! Tricycle is awesome!
A relative of mine is a huge Operations Research fanboy, but I don't get it - I just see graph theory and calculus, and some clever but brittle planning algorithms. If you can explain what the big deal is, please do.

"Religious issues" in hardware and software

Apparently, it's not just politics that is the mind-killer. When it comes to one's tool of choice, one can get as irrationally fanatical as it gets. From the Jargon Dictionary

“What is the best operating system (or editor, language, architecture, shell, mail reader, news reader)?”, “What about that Heinlein guy, eh?”, “What should we add to the new Jargon File?”

Great holy wars of the past have included ITS vs.: Unix, Unix vs.: VMS, BSD Unix vs.: System V, C vs.: Pascal, C vs.: FORTRAN, KDE vs, GNOME, v

... (read more)
My impression is that holy wars about software and hardware aren't as common as they used to be. Is this correct?

Apparently, hammocks are really good to sleep on.


Alvin Plantinga is renowned as one of the finest philosophers and theologians that the theist world has to offer. Maarten Boudry (the guy who did the Sokaling I note a couple of posts down) absolutely obliterates the painfully awful bloviations on science and evolution to be found in Plantinga's latest book. PDF, p21 on. If you enjoy LessWrong, you will enjoy this example of a good philosopher skewering bad philosophy.

Related: I have finally written up the recent history of the phrase "sophisticated theology" as used by New Atheists and their fan... (read more)

Just a few days ago it occurred to me that reductionism is the inevitable result of strictly local causality -- or putting it differently, any universe with a speed-limit in causality (in our universe this seems to be c) must by necessity be a reductionist universe.

The conclusion seems obvious but the connection between the two had never before occurred to me.

How so? In a local nonreductionist universe, wouldn't a fundamentally complicated thing just have to have zero width?
Ah, I guess I was thinking of a different type of non-reductionism, where e.g. an organism gained an extra non-reductionist property "life" which caused all its parts to behave differently than if they lacked that property...

Maybe this has been discussed here, but I wanted to see what you guys think of the surprise test paradox.

The paradox goes like this: it is impossible to give a surprise test.

Say a teacher tells her class on Monday that there will be a surprise test this week. The test cannot be on Friday, because when no test has been given by end-of-class Thrusday, everyone will know that the test is on Friday, and so it won't be a surprise. The test cannot be on Thursday. Having established that the test cannot be on Friday, if no test has been given by end-of-class on W... (read more)

Mixed Nash equilibrium is going on here :P If you made this into a little game, with the students paying a cost to cram but getting a benefit if they crammed the night before the test, and the teacher wanting the minimum number of students to cram, you could figure out what the actual ideal strategies would be, and the teacher would indeed have a mixed strategy. What the non-probabilistic (that is, deductive) reasoning really shows is that there is no way to always surprise a student. If you make things probabilistic, that means you're only claiming to surprise your students the most you can. This problem is weird because it demonstrates how in order to be surprising overall, sometimes you actually do have to choose bad options (at least when you're playing against perfect reasoners :D )! It's only by sometimes actually choosing Friday that the teacher can ever get students to not cram before class Thursday - the times the quiz is on Friday are sacrificed in order to be surprising the other times.
That's a great reply, thanks!
It's an amusing paradox and I don't know what the standard solution to it is, but I resolve it easily in my mind by tabooing the word 'surprise', and replacing with "suddenly obtaining certain knowledge of its date". Then it becomes more of a silly game of words. Assuming the test is certain to take place (it's the law!), the students can't be suprised on Friday, but on Thursday they will either be "surprised" to locate the test on Thursday or "surprised" to locate the test on Friday. The Thursday (or Wednesday or Tuesday or Monday) surprise will therefore be genuine, though no student can be suprised on Friday. They can be "surprised" on Thursday about either a Thursday or a Friday test, though. The day-by-day progression is a distraction btw. The paradox can be replaced by having five cups, a ball under one of them, lifting them one by one in whatever order. If you've lifted three empty cups, you'll then either be "surprised" to locate the ball under the cup you lift next, or "surprised" to locate it definitively under the cup you haven't lifted. Both of these will be a surprise, a surprise occurring at the next-to-last cup.
The 'no friday, but any other day is fine' thing is the closest we have to a standard solution. The taboo game is a good idea, but it could also easily be misleading. The taboo solution works or doesn't work depending on what you replace 'surprise' with, so you have to argue for your replacement. (EDIT: this strikes me now as a general and serious problem with the taboo game) Here's an argument against your replacement. If I told you that there would be a test next year, on this day, at exactly 2:00, you would hardly call this a surprise even though you'd just gained 'sudden and certain knowledge of its date'. It would be impossible to not give a surprise test. This can't be what 'surprise' is supposed to mean, and even if it is, the paradox still makes impossible another kind of surprise test which teachers often take themselves to be able to announce. A surprise test should probably be understood as a test given in such a way that you do not know, the night before (when you would study) that the test will be the next day. This is what the paradox makes problematic. The students can't be surprised, on Thursday, to locate the test on either Friday or Thursday, because they know the test won't be on Friday (since then they'd know about it the night before). So they know the test must be on Thursday (which they would have guessed last night, so no surprise there either). You can reproduce the paradox with five cups, yes. But the conditions have to be narrower than you say: the cups have to be lifted in a specific order known to the lifter before hand. The lifter has to be told that he will not know, before he at any stage lifts the cup to find the ball, wether or not the ball will be under that cup. So the player will know that the ball cannot be in the last cup (since then he will know before he lifts it that it is there), and given that, it cannot be under the second to last cup either, and so on.
Furthermore, if you take the "information content" approach to surprise, then you would be more surprised by a test on Monday than by a test on Thursday; but this is made up for by the fact that on Monday, Tuesday, and Wednesday, you were very slightly surprised that there wasn't a test. Total surprise is conserved.

My 7-year-old son likes computer programming and I suspect has a lot of innate aptitude for it. We have worked out a system that for every X minutes of learning he does with me he gets X minutes of computer gaming time. What kind of learning exercises could help him be a better programmer when he becomes an adult? Should I focus on him doing lots of coding or, for example, would he be better served by learning additional math?

For those of you who are adult computer programmers (and please identify yourself as such in your response) what, if anything, could you have done at a young age that you think would have caused you to now be a better programmer?


I'm a hobbyist computer programmer considering a career in it.

When I was 6, I met a friend who was into star-trek and science and such. We used to talk about science stuff and dig up "dinosaurs" and attempt to build spaceships. I think a lot of my intellectual personality came from being socialized by him. The rest came from my dad, who used to teach me things about electricity and physics and microeconomics (expected value and whatnot).

I learned to program when someone introduced it to me and I realized I could make video games. (I was 18) I absorbed a lot of knowledge quickly, and didn't get much done. I would find some little problem, and then go and absorb all the relevant knowledge I could to get it exactly right. Even tho I didn't accomplish much, now I know a lot about computer science, which is helpful. Having some thing I was trying to do put a powerful drive behind my learning, even if I didn't actually act in a strategic or effective way.

My dad occasonally told me the importance of finishing the last project before beginning the next, but I don't think it properly transferred. I still have lots of trouble shipping.

One thing that bit me a lot was regressing into... (read more)

Of all my flaws, I currently consider my bias to thought (and study, research, etc.) over action my greatest. I suspect that LessWrong attracts many (a disproportionate number of) such people.

I am a software developer, and have glimpsed over many similar questions. To summarize: There are enormous individual differences in how one can become a better programmer, and even more so on the opinions on it. It is not even easy to agree on what basic skills should be there at the "end" (i.e., the beginning, after your first two years real experience), much less on how to get those skills.

That said, most commonplace advice is valid here:

  • there is with high probability no really significant innate aptitude for programming (intelligence has a carry-over into many domains, however) (edit: seems I am wrong about that: See "Has “Not everyone can be a programmer” been studied?" at Programmers.StackExchange
  • don't expect too much skill carry-over: mathematical thinking and programming are related, but different activities
  • being able to concentrate for some time is good (a.k.a.: conscientiousness helps)
  • as a kid, liking the task probably also helps
  • playing the trumpet - not that much (though it can be fun)

On skill-set:

  • The practical coding skills are always important. (Except if you want to earn more than a dev.)
  • Algorithmic knowledge is rarely helpful, but you
... (read more)
One thing I have observed about people who are better programmers than me is that many play musical instruments. (I don't.) I have no idea what sort of causal structure is involved here. Learn off-sequence math: logic puzzles, geometric constructions, number theory (primes, modular arithmetic) that are not represented in the standard school math sequence. Learn something of electronics, digital logic, etc. Use different languages: maybe Python and Pygame for animations and games, and HTML and JavaScript for web pages. Don't allow the idea to develop that there is one "normal" way for code to look, e.g. "normal languages have curly braces; ones that don't are gratuitously weird."
I'm an adult professional programmer. The gaming machine of choice when I was a kid was C-64, not NES, so I knew programming your own stuff was a thing at elementary school age. The obvious interesting thing with the computer was game programming. The (generally correct back then) received wisdom also was that any serious game programming effort with the 8-bit home computers involved assembly languages, but I wasn't able to find any manuals or software I could understand to teach myself assembly, so I never ended up learning it. A few years later the PC was the top gaming box, and assembly had given way to C (and Turbo Pascal, which was a big thing on DOS back then for some reason), I managed to learn C after a false start or two, and went off happily coding many stupid things I did not finish with it. So not much of a lesson here, not learning assembly at age 10 didn't seem to slow me down much later, and it also turned out that I never did need to learn to program 8-bit microcomputers right down to metal to make games. Still, I'm sure it would've been interesting to learn that. The big thing that made me learn programming was the intrinsic drive to want to program computer games. Any sort of achievement-endorsing authority figures were generally entirely indifferent to this. The hard things in programming for me are also ones which I suspect are inherently boring. Working with legacy codebases, working on uninteresting features that are needed for business reasons, finishing the countless little things like documentation, ports and packaging so that something can be released, and in general going from the 80 % solution to the 100 % one. I guess I could echo nyan_sandwich's admonition to teach shipping here. I did waste some time writing unnecessarily bad C++ before I read Stroustrup's book on the language, so I guess hitting the books early could help. I also crashed, burned and never recovered enough to have any degree of academic success with university leve
I'm a computer programmer who started at age 12. At that time, by far the coolest thing about programming for me was being able to write my own game. I was also motivated by math and algorithms, but more weakly, at least initially. I would recommend balancing two things. First, settle on a simple framework that lets you do simple graphics in a simple language with minimum boilerplate. Maybe pygame (never tried it), or processing.js (good because it works right there in your browser), or some learning-oriented language (when I was a kid, Logo was popular, though I quickly moved on to BASIC; is there an analog of Logo for the 2010s?). Show him how he can draw something with three lines of code, get him fascinated, move the goalposts, iterate. Second, more math but I would especially advise math puzzles of all kinds. Tricks with numbers, geometric constructions, logical puzzles. It's great if you can find material of this kind for your son that he'll want to consume on his own and not just because of the time-sharing scheme. Martin Gardner's books may be useful, though probably a bit later on; and other material of such kind. Note that I haven't answered your last question because while the things I listed were helpful to me, I did them, and I don't know what other, different things would've caused me to be a better programmer. I don't even know if starting at age 7 would've caused that; I suspect it's true, but don't have a solid argument.
You might try getting him a shell account and have him learn to use tmux (or screen) and vim (or emacs).

I'm thinking of starting a meetup group in Auckland, NZ, but would like to gauge the interest in such a group first. I know I'm supposed to just plan a meetup, but from a search of the site it looks like Auckland groups have failed to get momentum in the past, so I'd like to arrange for a time and place such that at least a couple of people can definitely attend.

Reply if you're interested, and then we'll sort something out.

I think gauging interest for meetups is usually presented as a discussion post so that it's more likely to be noticed, but I'm not sure I'm right about that. Anyone else notice a pattern?
Thanks. I thought that that might be the case, but I wasn't sure either. If I don't have anyone expressing interest here by Sunday 6pm local time I'll make a discussion post, and if I don't get any replies to that in a day or two then I'll just add a meetup.

Most economists seem to agree that the costs of voting outweigh the benefits. I've been considering whether it might be worth trying to use the threat of voting to get people to do something costly (from their perspective) which serves my values. Here is how I would picture it working:

  1. I commit to vote for a candidate (if my conditions are not met). Contrariwise I commit to
  2. People who dislike that candidate will want me to abstain rather than casting my vote.
  3. Require something which seems non-selfish so as to avoid sacred value dissonance.

So for exampl... (read more)

How does ambient decision theory work with PA which has a single standard model?

It looks for statements of the form Myself()=C => Universe()=U

(Myself()=C), and (Universe()=U) should each have no free variables. This means that within a single model, their values should be constant. Thus such statements of implication establish no relationship between your action and the universe's utility, it is simply a boolean function of those two constant values.

What am I missing?

The problem is that the agent doesn't know what Myself() evaluates to, so it's not capable of finding an explicitly specified function whose domain is a one-point set with single element Myself() and whose value on that element is Universe(). This function exists, but the agent can't construct it in an explicit enough form to use in decision-making. Let's work with the graph of this function, which can be seen as a subset of NxN and includes a single point (Myself(), Universe()). Instead, the agent works with an extension of that function to the domain that includes all possible actions, and not just the actual one. The graph of this extension includes a point (A, U) for each statement of the form [Myself()=C => Universe()=U] that the agent managed to prove, where A and U are explicit constants. This graph, if collected for all possible actions, is guaranteed to contain the impossible-to-locate point (Myself(), Universe()), but also contains other points. The bigger graph can then be used as a tool for the study of the elusive (Myself(), Universe()), as the graph is in a known relationship with that point, and unlike that point it's available in a sufficiently explicit form (so you can take its argmax and actually act on it). (Finding other methods of studying (Myself(), Universe()) seems to be an important problem.)
I think I have a better understanding now. For every statement S and for every action A, except the A Myself() actually returns, PA will contain a theorem of the form (Myself()=A) => S because falsehood implies anything. Unless Myself() doesn't halt, in which case the value of Myself() can be undecidable in PA and Myself's theorem prover wont find anything, consistent with the fact that Myself() doesn't halt. I will assume Myself() is also filtering theorems by making sure Universe() has some minimum utility in the consequent. If Myself() halts, then if the first theorem it finds has a false consequent PA would be inconsistent (because Myself() will return A, proving the antecedent true, proving the consequent true). I guess if this would have happened, then Myself() will be undecidable in PA. If Myself() halts and the first theorem it finds has a true consequent then all is good with the world and we successfully made a good decision. Whether or not ambient decision theory works on a particular problem seems to depend on the ordering of theorems it looks at. I don't see any reason to expect this ordering to be favorable.

After reading comments to a certain work of fiction published here I was surprised to discover strong negative reaction to smoking of the protagonist. (personally, I found this part a nice touch as far as its literary qualities) Thus for my fiction-writing purposes I would like to know:

What is the general stance on drug usage in fiction?

  1. Smoking marijuana

  2. Using psychedelic substances

  3. Using unidentified or fictional substances of supposedly psychoactive nature, but with their full effect or social status unmentioned.

I do not think there is a "general stance" on drug usage. Further, tobacco is not a general drug (there's no such thing) but a specific well-known drug with a specific history. I would not expect reactions to tobacco smoking in fiction to be reflections of general stances on drug use. ---------------------------------------- What I read in the reaction to Hanna's smoking is specifically that, today and to some readers, smoking seems ① self-destructive and ② odious; and furthermore ③ signals unsympathetic or disfavored character tropes. First, because we know so well today that tobacco smoking causes such a wide range of deadly diseases, an educated character in present-day or near-future fiction can be assumed to know this, too. She has willingly taken up a habit which causes cancer, emphysema, heart disease, etc. Many readers will have had relatives who died due to smoking-related diseases. So smoking may be seen as a symbol of a character's self-destructiveness. (An example that comes to mind is John Constantine in the Swamp Thing and Hellblazer comics.) Second, smoking in public is often seen as odious — or, at least, insensitive to others' well-being. People from social backgrounds where smoking is uncommon may, due purely to a selection effect, associate smoking with people who do not care that they are being unpleasant to others. So smoking may be a symbol of a character's rudeness, hostility, or willful repulsiveness — a choice to "gross out" others to assert personal space, for instance. (Also, many readers may have sensitivity to smoke, asthma, allergies, or simply a strong dislike for being around smoking or people who smell of smoke. So they may read a character smoking and think, "I would not want to be around her," and through mind projection fallacy conclude "she is unpleasant to be around.") Third, smoking may be culturally associated with various sorts of characters (or real-world people) whom the readers disfavor — beatnik-type hipsters; redne

New post idea:

"Female Test Subject - Convince Me To Get Cryo"

(Offering myself for experimentation, of course.)

What do you think? Should I post it? [pollid:110]

Sounds like a good idea to me. Cryonics could definitely use more female-audience-targeted arguments.
What's the point?
To practice on me before something happens to your female family members and you've got to convince them... Friendly hint: you just implied my life isn't worth saving. I am not easily offended and I'm not hurt, so that's just FYI.
Are you such a Platonically ideal female that we can generalize from you to other females, who may have expressed no interest in cryonics? If you see it that way, it sounds like you're already very nearly convinced.
Of course not, that's an assumed "no". I guess what you're really asking is "What is the point of seeing whether we can convince you to sign up for cryo?" Sometimes case studies are helpful for figuring out what's going on. Study results are more practically useful but let's not forget how we develop the questions for a study - by observing life. If you've ever felt uncomfortable about the idea of persuading someone of something or probing into their motivations, you can see why being invited to do so would be an opportunity to try things you normally wouldn't and explore my objections in ways that you may normally keep off-limits. Even if most of my objections are different from the ones other people have, discovering even a few new objections and coming up with even a few new arguments that work on others would be worthwhile if you intend to convince other people in the future, no? Alicorn is right. It's not that I am convinced or not convinced, it's that I'm capable of interpreting it the way that you might have meant it. For the record, where I'm at right now is that I'm not convinced it's a good way to save my life, (being the only way does not make it a good way) and I'm not 100% convinced that it's better than donating to a life-saving charity.
I'm trying to say that I think you might already be a pretty extreme outlier in your opinion of cryonics, based on a few clues I noticed in your comment, so your reactions may not generalize much. The median reaction to cryonics seems to be disgust and anger, rather than just not being convinced. I'm sort of on the fence about it myself, although I will try to refute bad cryonics-related arguments when I see them, so on object-level grounds I can't really say whether convincing you or learning how to convince people in general is a good idea or not.
Disgust and anger, that's interesting. I wonder if that might be due to them feeling it's unfair that some people might survive when everyone else has died, or seeing it as some kind of insult to their religion like trying to evade hell (with the implication that you won't be motivated enough to avoid sinning, for instance). If that's the case, you're probably right that my current reaction is different from the ones that others would have. My initial reaction was pretty similar, though. My introduction to cryo was in a cartoon as a child - the bad guys were freezing themselves and using the blood of children to live forever. I felt it was terrifying and horribly unfair that the bad guys could live forever and creepy that there were so many frozen dead bodies. I didn't think about getting it myself until I met someone who had signed up. My reaction was "Oh, you can actually do that? I had no idea." - and it felt weird because it seemed strange to believe that freezing yourself is going to save your life (I didn't think technology was that far along yet), but I'm OK with entertaining weird ideas, so I was pretty neutral. I thought about whether I should do it, but I wasn't in a financial position to take on new bills at the time, so I stored that knowledge for later. Then, when I joined LessWrong, I began seeing mentions of cryo all over. I had the strong sense that it would be wrong to spend so much on a small chance of saving my own life when others are currently dying, but that was countered pretty decently by one of the posts linked to above. Now I'm discovering cached religious thoughts (I thought I removed them all. These are so insidious!) and am wondering if I will wake up as some sort of miserable medical Frankenstein. I can't tell you whether it's worth it to convince me or learn to convince people, either. I'm not even sure it's worth signing up, after all. (:
She could know that you see it that way without seeing it that way herself. If I knew someone who believed that I would definitely go to hell unless I converted to their religion, and they didn't seem to care if I did that or not, I might characterize that as them not thinking my soul was worth saving.
Yeah, that's true. But still, if "they don't think my soul is worth saving" is more salient to you than, for instance, "I'm glad I won't have to deal with their proselytizing," it suggests that you take the idea of souls and hell at least a little bit seriously. To give a more straightforward example, imagine a police officer asking someone someone whether they have any contraband. The person replies, "no, officer, I don't have any weed in my pocket." How would that affect your belief about what's in their pocket?

Dual N-Back survey

Please take it if you've done n-backing regularly at some point!

on hpmor: harry seems to be very manipulative but almost in a textbook kind of way. I take it eliezer got this from somewhere but cannot figure out where. I'd love to read more about this, could it have come from "the strategy of conflict"?


I haven't read Strategy of Conflict, but I have read Robert Cialdini's book Influence: The Psychology of Persuasion, which Harry name-drops a couple of times and uses several techniques from. I'd guess that that's some of what you're seeing. For future reference, though, it's considered polite to confine free-floating HPMoR discussion to the Methods threads, the most recent of which appears to be here. There have been a few Methods-related threads since, but all with narrower scope.
Thank you. I had read the Methods threads but these seemed mostly to discuss the plot of the books. Anyways, message received, will update accordingly.

A talk on The Importance of Mathematics by Timothy Gowers