Log in

No account? Create an account
thomascolthurst's Journal
[Most Recent Entries] [Calendar View] [Friends]

Below are the 20 most recent journal entries recorded in thomascolthurst's LiveJournal:

[ << Previous 20 ]
Saturday, February 19th, 2011
9:44 pm
The Alphabet of Suburbs!
 (With apologies to TMBG)

Arlington, Belmont, Cambridge, Dedham, Everett, Framingham, Gloucester,
Haverhill, Ipswich, Jamaica Plain, Kingston, and Lexington.
Medford, Newton, Oxford, Peabody, Quincy, Revere, Somerville
Taunton, Uxbridge, Vallersville, Watertown, eX-yankees fan, Yarmouth, Zylonite!
Thursday, February 17th, 2011
4:00 pm
Weird Al's "I Lost on Jeopardy" video is still worth watching.
I worked in the natural language processing field for ten years, and even at one point knew some of the folks doing the same at IBM research. Plus, I hate humans, so naturally I was rooting for Watson during Monday and Tuesday's shows.

But Ken Jennings earned my support with just a single word on Wednesday. (And no, that word wasn't "fity", though I did appreciate his fluent pronunciation of "Who is 50 Cent?".)

The word was "VillageReach", which just happens to be the name of the best charity on the planet.

I'm not being intentionally hyperbolic when I say that, and I don't mean "best in my subjective opinion". I really and truly mean that VillageReach is objectively the best organization you can give your money to if your goal is helping people.

But don't take my word for it. Take givewell.org's. They evaluate charities based on impact; i.e., how the charity effects the people it touches. (You would think this would be the obvious and universally used method of evaluating charities, but you would be surprised.)

Anyway, VillageReach is givewell.org's top rated international charity. VillageReach does unsexy logistical work in delivering medical supplies to rural Africa, but their cost effectiveness is astounding, at approximately $600 per infant life saved (+/- $200). (It is important to emphasize that this is tested and measured effectiveness, based on the work that VillageReach has already done, as opposed to a guess from an academic paper.)

Let me flesh that out for you: with the $150,000 Ken is donating from the Jeopardy contest (half of his $300,000), VillageReach will save approximately 250 lives. Had Ken won the $1 million top prize instead, almost (and again, approximately) 600 additional lives would have been saved.

To be fair, IBM is donating Watson's $1 million prize to World Vision and the World Community Grid. I don't know whether those are good or bad charities, but my point is that neither does anyone else: neither charity has ever attempted to evaluate their impact.

Okay, end of rant. For those of you still with me who are interested in how Watson did its stuff, I recommend this AI Magazine paper.
Friday, October 29th, 2010
7:55 pm
Blogs I'd Like to See #4
Definitions From the Future.  Every day you get a definition of a word or phrase that doesn't exist yet, but will.  Examples:
  • Antitwin:  two people are antitwins if they are siblings who share no DNA, other than the DNA that their father and mother have in common.  Because of the mechanics of crossover, your antitwins are genetically unique in the same way that your twins are.  A true antitwin does not even share mitochondrial DNA (i.e., gets their mitochondrial DNA if you like most people get yours from your mother); some pedantic sticklers take the word antitwin to mean "true antitwin".  Yes, there are still pedantic sticklers in the future.  Antitwin can be used as either a noun or a verb:  "Sheila was sad that Bobby didn't get her eye color, so we antitwinned him to get Alicia."
  • Adaptive Food (AF):  food that reconfigures its molecularly surfaces based on information derived from the mouth environment of the person consuming it.  Typically the reconfiguration aims to achieve the same subjective taste despite the wide variance in the numbers, types and placements of taste buds on human tongues.  (See, for example, supertasters).  For economic reasons, most AF food contains only a small number of mass produced AF ingredients, such as AF salt or AF corn syrup, but in New High Cuisine it is not uncommon for an entire meal to be AFized.
  • Uplift Pet Therapist.  You love your dog Fido, so you went ahead and paid for his sentience operation.  Well, guess what?  Fido has lots of issues with his newly enhanced consciousness, and in the new EU at least, you are legally obligated to pay for two years of his weekly therapy sessions.  There is no such obligation in the old USA, but it is still probably a good idea, because it turns out that canis lupus sapiens loves to blog ...
Friday, September 10th, 2010
7:19 pm
Blogs I'd Like To See #3
 Highbrow/lowbrow.  As the name suggests, the posts on this blog would strictly alternate between high and low culture.  Example sequence:
  • John Adams is a modern day Bernstein, an ultimately minor composer wildly overrated simply by virtue of being the best among Americans. Besides On the Transmigration of Souls, what works of his will be listened to fifty years hence?  And don't say Nixon in China -- Donal Henahan got it right when he reviewed that work, saying "Mr. Adams does for the arpeggio what McDonald's did for the hamburger."
  • Sometimes after a good #2, I like to sing, "Poop!  There it is!  Poop!  There it is!"
  • When one thinks of Susan Sontag, "philistine" is perhaps the last word that comes to mind, but consider this:  when she and Black Swan author Nassim Taleb met at a BBC studio, Sontag initially was interested in talking to Taleb up to the moment she discovered that he worked as a market trader, at which point Taleb reports that "She turned her back to me as I was in mid-sentence."  It is hard to reconcile that sort of close-mindedness with the roving intellect behind Against Interpretation, but I suppose we all contain multitudes.
  • [embedded youtube video showing a toddler pushing over two cats]
Monday, September 6th, 2010
9:00 pm
Wanted: funny fake game titles
So it looks like Barons! is actually going to get printed this year, as I just got an email from the publisher asking for a paragraph long bio.

Said bio is allowed to be fictional, which means that it is mandatory that I stuff it full of pretend games which I will claim to have previously designed.

Here's what I've got so far:

1970:  The Battle for Kent State
Snakes & Lagers, A Child's First Drinking Game
Par Cheesy  (The Game of Golf Puns)
Spleen!  The Wacky Organ Donor Game

Your suggestions are required.
Friday, September 3rd, 2010
4:46 pm
 Various folks have asked me how I liked my month as a visiting fellow at the Singularity Institute, and since telling them one at a time clearly doesn't scale, I thought I would bore everyone about what I did on my summer non-vacation.

Also, because I am a geek, it is in FAQ format.

Q: Did you like it?
A: Yes.

Q: What is the Singularity Institute?
A: The Singularity Institute (SI) is a non-profit think tank; their main concerns are the long term risks of human level and beyond human level artificial intelligence. In practice, what they mostly do is write papers about these risks and what we might do about them. Along with Ray Kurzweil, they also run the Singularity Summit, a yearly conference which despite its name is more of a grab bag of lectures a transhumanist might be interested in than in the singularity per se. The Singularity Institute has no formal ties with the Google affiliated Singularity University, but various SI folks have given talks there.

Q: What the heck does the word "singularity" mean?
A: There are two widely used but different meanings. The Kurzweil definition is an asymptote of accelerating exponential technological progress, beyond which prediction is more or less impossible. The older definition (which the SI uses) is of an "intelligence explosion": an artificial intelligence that is smart enough to recursively make itself smarter, with no diminishing returns up to the laws of physics.

Q: Is the singularity near?
A: No one knows. Some super genius might figure out the nature of intelligence and program one next week, or it might never happen. If you go to http://theuncertainfuture.com/index.php, you can generate your own probability distribution over years it might occur, based on your estimates of individual factors.

Q: Isn't it way too early to be worrying about the singularity?
A: For a given individual, maybe. For society as a whole, almost certainly not. The SI position is along the lines of: wouldn't it have been nice if folks had thought long and hard about the dangers of nuclear weapons decades if not centuries before we had created them?

Q: What did you do at SI?
A: A bunch of different stuff. SI is what I call a "startup think-tank"; they are still pretty new and don't have a lot of funding or permanent staff. A good portion of the time I just chipped in with whatever needed to be done that week, day, or hour. Among other things, I helped someone to learn how to program in C, I helped write an NSF grant application, I helped prepare for the Singularity Summit, I proofread a bunch of stuff that other folks were writing, and I wrote a rough draft of a philosophy/CS paper applying Goodhart's Law to AI.

Q: What the hell is Goodhart's Law and what does it have to do with AI?
A: Goodhart's Law states that "Any observed statistical regularity will tend to collapse once pressure is placed upon it for control purposes." It comes from economics, where "control" means "policy." An example would be the Federal Reserve observing a strong correlation between inflation and unemployment, a correlation which breaks down the second the Fed tries to manipulate the inflation rate to reduce unemployment. The paper argues that AIs have to deal with a variant of Goodhart's Law, especially as they grow more powerful.

Q: How was the Singularity Summit?
A: Fun! My favorite talks were by my life long hero James "The Amazing" Randi, and by Dr. Lance Becker, who I had never heard of before the summit, but is doing kick ass things in ER rooms with artificial circulation plus controlled reperfusion. And by "kick ass", I mean raising the dead: at one point Dr. Becker talked about a study which started with six legally dead people who post-treatment "had a 50% survival rate".

Q: What were the most important things you learnt over your month at SI?
A: (1) AIXI is the real deal, a fundamental advance in AI theory on the order of say minimum description length. For some of you, that was already obvious, but as a mostly AI-outsider I had previously not recognized its importance.
(2) Climate change models are extremely important, but not because of global warming. Even extreme global warming would with high probability take over a century to kill all of us, but no one right now knows the chances of a small scale nuclear exchange (between India and Pakistan, say) causing agriculture to not work world wide for over a decade. Or rather, current climate change models say that that is a likely possibility, but I would really feel more comfortable with that conclusion if the models were open source and not written in spaghetti code Fortran.

Q: What was the worse part of your month?
A: Sourdough bread. That stuff is a blight over the entire greater SF area.
Tuesday, August 17th, 2010
2:48 pm
Michael Kinsley & Climategate
The recent news about data sharing leading to progress in fighting Alzheimer’s reminded me that I wanted to make one point about the long dead and forgotten about "Climategate" scandal.

The point isn't even mine, it's Michael Kinsley's, namely his often used dictum that "the scandal isn't what's illegal; it's what's legal."

Which is to say:   I agree with the three independent reviews that the Climate Research Unit and Phil Jones acted within the normally accepted range of contemporary scientific practice.  The issue is that I and many others find that range, well, scandalous.

The scandal has many dimensions, but the specific one I want to talk about here is the availability of data and computer code.  My argument is simple:  science depends on the ability of researchers to reproduce, test, and extend results found by others, and that is impossible when the result is based upon data and computer code that is not made freely available.

And to fill out the minor premises:  science is good, or more specifically, results that have been exposed to reproduction, testing, etc. are orders of magnitude more reliable than results that have not been so exposed, and that (1) encouraging such reliability is good and (2) distinguishing between information sources with vastly different degrees of reliability is also good.

I'm not just advocating open data for science.  What I propose is:
  1. The NSF, NIH, DARPA and other government funders of science require that any projects or researchers they fund make all data and computer programs they develop as part of their funding available without cost and without encumbrance for future research.  I'm fine with people who don't like that restriction applying to the NEA.
  2. Science journals do not publish papers unless the data and computer programs used in the production of the research are available for a reasonable cost.  (It's okay, for example, to use Mathematica, Mac OS, or other such commercial programs.)  Scientists who dislike this restriction are still free to blog, tweet, and issue press releases about their research, or publish such in People magazine.
  3. Universities and colleges only make the research component of tenure and other hiring decisions according to papers meeting the above criteria.  I don't want to discourage scientists from blogging, but I would categorize that as an educational, not research activity.  (And yes, universities and colleges should weight educational activities much more than they do now when making tenure decisions.)
  4. Public policy, particularly policy affecting billions of people and touching trillions of dollars of economic activity, be based to the maximal extent possible on papers meeting the above criteria.
In short, I propose that we take back the word science, and reserve it only for research that is openly reproducible and testable.  We will of course need a name for the other stuff that we used to call science, something like "non-creative writing".

Now, the strongest and most obvious counter argument to all of this is that the benefits of such open reproducibility would be outweighed by the costs in reduced information flow.  We want, for example, the world to know that high temperature superconductivity is possible as soon as possible (so that post-docs and the like can start preparing to move into the field), but the under the proposed scheme they might be tempted to sit on their data until they have squeezed the last drop of analysis from it, lest they be scooped by another group.  I think the issue here can be solved, if indeed it needs solving, be slightly changing the incentives around press releases.  I think something like a press release is what we want here -- something that communicates the news, but also communicates its provisional nature -- and while there clearly are some incentives to issue press releases now, those incentives may have to be increased.

Another potential casualty would be research about, or that depended upon, expensive proprietary computer programs.  I used to this sort of research myself, in the field of speech recognition.  There were (and to my knowledge still are) no good open source speech recognizers, so the majority of research was done by the ten places that had a state of the art recognizer, and none of it was reproducible unless you were at one of those ten places.  And even then, it was darn hard (and sometimes even impossible) to reproduce another group's result, because no one shared source code.  For example, sometimes the alleged improvement was really just (unintentionally) covering over a bug in the other guy's code.  I shed no tears at the thought of not calling this sort of thing science.

Phil Willis, the chairman of the House of Common's investigative committee into Climategate, described the "standard practice" of not routinely making the data and programs used in climate science as "reprehensible", saying it "needs to change and it needs to change quickly".  He's right, and he's right even with "climate science" replaced with just "science".  If I had Alzheimer's, I would be happy that the researchers in that field had started sharing data and I would be furious that they hadn't started doing so twenty years ago.
Thursday, July 22nd, 2010
1:07 pm
Blogs I'd Like to See, #2
A Philosophical Skeptic Reads the News. Every day this blog takes a different news story and critically examines its philosophical assumptions.  Sample post:
"As we've come to expect from the New York Times, Toyota Receives Subpoena on Steering Flaws once again presumes -- without comment, argument or proof -- the existence of an external world. Worse, the problem of other minds is entirely elided over, to the point of blithely buying into what Churchland so rightly derided as folk psychology."
1:04 pm
Blogs I'd Like To See, #1
The Meme's The Thing:  The Royal Shakespeare Company re-enacts popular Internet memes, Bard-style.  Sample videos: Sonnet 73 on Treadmills, The Evolution of Early to Late Elizabethan Dance, Dancing Hamlet, Diet Coke and Juliet, Macbeth Kid, All Your Henrys Are Belong to Us, and Two Gentlemen of Verona One Cup.
Wednesday, May 12th, 2010
5:51 pm
Continuum Arguments
A continuum argument purports to show that two things are the same by presenting a continuous sequence of intermediate states between the two. For example, if I were to argue that black and white were the same color, because there is clearly a way to morph white into black through a movie of grays, each frame imperceptibly different from the next, then I would be making a continuum argument.

As the example suggests, continuum arguments are almost always invalid. Indeed, Wikipedia files them under continuum falacy. I think that is a little bit harsh, because there are at least two technical areas in which continuum arguments hold force:

1) In the mathematical field of topology, two shapes are explicitly considered to be the same if they can smoothly be changed into each other. As the standard joke goes, a topologist can't tell her donut from her coffee cup. In topology, then, continuum arguments are far from falacious; rather, they are proofs by construction.

2) In the philosophical field of ontology, which studies the fundamental nature of existence. Here the standard joke is that at least in English, the entire field of ontology can be summarized in a two word question and one word answer: "What exists? Everything."

Continuum arguments are powerful in ontology because of the requirement that the fundamental entities actually be fundamental, and not themselves made up of other different stuff. For example, let's say you are a Cartesian dualist who believes that mind and matter are two totally separate and essentially different things. Then if I can show a continuum between something which clearly has a mind (Stephen Hawking, say) and something that does not (Hawking's great^(10^10) amoeba-like grandfather), you are in trouble. It isn't a completely knock-down argument: you can retreat into panpsychism, or say that the amount of mind slowly or abruptly decreases as you move along the continuum (but why would it do that if mind and matter were completely different?). But it does move you from a simple "common sense" dualism to a more complicated and much less attractive version.

Continuum arguments are sometimes confused with slippery slope arguments, but they are quite different, as a slippery slope is causal -- "if you allow A, then B will also happen, and B is bad" -- and a continuum argument isn't. Slippery slope arguments are also not out and out fallacies like continuum arguments are, but neither are they in my experience persuasive: if you are arguing about A, then the claim that A leads to bad thing B is likely to be just as contentious as A itself. Or more so, given that predicting the consequences of nearly anything is known to be hard, at least outside of physics where there really are nearly frictionless inclined planes ...

The recursively minded among you will now be wondering whether there are any fallacies intermediate between a continuum argument and slippery slope. I'm happy to say that there is. It doesn't have a Wikipedia page (yet), but a "no bright-line" argument tries to show that A is bad by placing it on a continuum with known bad thing B. For example: one shouldn't go to see Iron Man 2 for $5 per ticket, because one shouldn't go to see Iron Man 2 for $50,000 per ticket. The same fallacy is at work in the famous "Now are just haggling over the price" Winston Churchill joke. It's not a slippery slope because there is no causality, and it isn't a continuum argument because it isn't trying to establish identity, merely morality or prudence. The name comes from law, in which a bright-line rule is a clearly defined function of objective factors: you must be 21 to drink is a bright-line test, while "I know when I see it" isn't. Here, the presence of a bright line would put a kink in the continuum -- if there was a well known rule of thumb that one should never pay more than $100 for a movie ticket, then the difference between $5 and $50,000 would be even more readily apparent than it already is. But while being bright-line may be a desirable property for a rule to have, there is no law of nature saying that all goods things are separated from all bad things by such; the mistake of a "no bright-line" argument is to assume that there is.

(Portions of this blog post were funded by Paramount Pictures, makers of Iron Man 2, available in theaters now.)
Sunday, April 18th, 2010
10:06 pm
The ten books which influenced me the most
Aka the meme Tyler Cowen kickstarted:

0. Star Wars: A New Hope. Not a book, but it influenced more than anything on the rest of this list. I saw it several times when I was five years old, and I've been in a science fiction frame of mind -- by which I mean a sense that things could be different, that anything was possible -- ever since.

1. Advanced Dungeons & Dragons: Dungeons Master's Guide. The first edition, Demon on the cover version that Gary Gygax wrote. This may be the first book I read that was way above my grade level; I think it was aimed at college kids while I was maybe eight when I first read it. There is a temptation in such situations to believe that the glimpse of the adult world you get is much more representative than it really is; all the more so if what you are reading is a book on world building. More than anything else, this book taught me to view the world generatively: whenever I learn a new fact, there is still some part of my brain that thinks, "How can I use this to be a better DM?" (Which these days another part of my brain translates into "How can I use this to sample from the space of possible worlds more accurately?") On the negative side, my writing style never really recovered from the exposure to all those Gygax-isms (i.e. and e.g. most especially).

2. Commodore 64 Programmer's Reference Guide. This first hard thing that I learnt entirely on my own was 6502 assembly language, and I learnt it from this book.

3. The Closing of the American Mind. These days I have no patience at all for the sort of cultural criticism done in this book, or really for any kind of cultural criticism at all come to think of it. However, this book was undoubtedly influential on me when I read it back in high school (it came out in 87, but I bought it on discount at Crown Books many months after that), in that it caused me to read a lot of classical philosophy.

4. Godel, Escher, Bach. I liked this book, and definitely went through a period where I saw recursion in everything. But it is on this list mainly for the other, better books it caused me to read: Metamagical Themas, for one, but also The Mind's I, which then caused me to read all of Daniel Dennett's books. Which are all awesome.

5. Chaos: Making of a New Science. The prologue of Chaos describes Mitchell Feigenbaum at Los Alamos, experimenting with 26 hour sleep schedules, working and waking whenever he wanted, getting paid to think deep thoughts. I read it and instantly knew that that is what I wanted to do when I grew up. This book also introduced me to Barnsley's "Chaos Game" algorithm for generating a Sierpinski Gasket, which led me to very successful science fair project and the joys of mathematical research.

6. Michael Barnsley, Fractals Everywhere. This is the book that I know better than any other book. (Marsden & Tromba's Vector Calculus comes in second.) I know the answers to all the exercises, I read all of the references, I can tell you how many pages each section is. I can still remember buying this from the UC Berkeley bookstore while visiting the campus my senior year of high school; it cost what seemed at the time to be the ungodly amount of $40. Of my own money, even!

7. The Mathematical Experience. My high school math teacher Doug Hazlett gave me this book as a graduation present, and while it would be overstating the case to say that it made me want to become a mathematician -- at the time, my preference was for cognitive science -- it definitely planted the seed in my mind. Which is slightly ironic, given that I'm fairly sure one of Philip Davis's aims for this book was to demystify and deglamorize what mathematicians do.

8. Free to Choose, Milton & Rose Friedman. Believe it or not, my political views were mostly liberal through high school. For the 1988 Constitutional Competition, for example, I remember defending a quite expansive reading of congressional and judicial power, as long as they were used "in the public good". (I was always something of a contrarian, though, and pretended to support Reagan in 84 so as to horrify my friends.) Reading Rand didn't make me a libertrarian, and neither did reading Nozick. (That said, I am in the small minority who thinks Anarchy, State and Utopia is a better piece of political philosophy than Rawl's Theory Of Justice.) This book's calm, relentless and empirical arguments did.

9. Richard Posner, Sex & Reason. A revelation; it changed the way I think about sex and human behaviour in general.

10. Judith Harris, The Nurture Assumption. I'd probably parent in mostly the same style that I do now had I not read this book, but I'm grateful for whatever the extent that it has shifted me away from trying to mold my child and towards simple kindness.
1:26 pm
Some thoughts on Inglourious Basterds
Warning: contains spoilers and aesthetic criticism.

I was thinking the other day about Inglourious Basterds. It's that sort of movie; it sticks in your mind (or my mind, at least) and comes up now and again, and not just when you are watching some other old movie and realize to your delight that *that's* where Tarantino stole that shot from.

One thing I haven't seen other reviews or analysis talk about is the extent to which Inglourious Basterds is a movie about the limitations of movies. Consider:

* Lt. Archie Hicox seems like a perfect choice for a spy because he is a British film critic specializing in German cinema. Alas, SS Major Dieter Hellstrom notes his odd accent and non-German hand sign for ordering drinks. Had Hicox's knowledge been less superficial, he might have lived.

* In the finale, Shosanna (aka Emmanuelle Mimieux) and Frederick Zoller have a gun fight within the hearing range of hundreds of people who would have surely heard them had they not been watching a war film at the time. Had the audience not been distracted, they might have escaped from the death trap that Shosanna created for them.

* Some people have complained about the decidedly non-historical ending of Basterds on the grounds that it is fantasy. Others have defended it as alternate history akin to Philip K. Dick's The Man in the High Castle. I don't think that's right -- alternate history starts with the change and lets the what-if play from there, while Basterds has the major ahistoricism at almost the very end and isn't at all interested in its consequences. Instead, I think Tarantino is moving up a meta-level to give yet another reason not to trust movies: they lie.

It's interesting to contrast these ways that movies lead us astray (they give us incorrect or superficial knowledge, they distract us from what's important) with another of Basterds's themes, that of the power of empathy. I'm using the word empathy in the limited sense of being to able to adopt another person's perspective, without necessarily sympathizing or sharing the emotions of that person. The character who has this quality in spades is Colonel Hans Landa, who we see using it to get inside and mess with the heads of those he interrogates; he also at one point explicitly credits his success as "The Jew Hunter" to his ability to think like a Jew.

It is striking that none of the nominal heroes of the movie (Shosanna, any of the Basterds) displays this sort of empathy; in fact, the only other character in the movie that I can recall demonstrating it is the German officer in the tavern scene who imaginatively compares the story of King Kong to the history of African slaves in America. (Not all the German officers are given empathy, though; Zoller in particular lacks any understanding of why a French woman might not want to date him.) Some of this is no doubt just Tarantino going in the opposite direction of audience expectations, but I think it is also a demonstration that empathy in and of itself is amoral, and in fact can be used for horrifyingly evil ends.
12:01 pm
How to Get Rich Fast
K. W. and I were talking the other day about "Hot Tub Time Machine" and that brought up the general topic of movies whose premise is contained in their title, "Snakes on a Plane" being perhaps the phenomena's epitome.  This caused me to wonder about films that betray the explicit promise of their name -- would any screen writer dare attempt such a thing?

Luckily, K. W. has an encyclopedic knowledge of cinema and was able to produce the example of Abbott and Costello Go To Mars, in which (spoiler alert!) Abbott and Costello do not go to Mars.  Since then, I've also thought of the following semi-examples:

Jurassic Park -- The park and movie are entirely located in the Holocene.

Star Wars -- Only one war is depicted, and though a planet is destroyed in the course of combat, the fight is between two bands of humanoids, with nary a massive ball of plasma expressing support for either side.

Citizen Kane -- Charles Foster Kane's legal status as an American is never proven, though I'm told there is a director's cut with a lovely tracking shot of his birth certificate that the studio made him remove.

A Clockwork Orange -- Do I need continue?

Anyway, I'm deliberately setting a low bar here so that you can do better in the comments.  You're welcome.

Tuesday, March 23rd, 2010
10:11 am
I really enjoyed Andrew Gelman's recent review essay on the relationship between causal inference and statistics. That might sound boring, but really it is the very heart of science: what can data tell us about causes?

As you might imagine, answers to that question vary a lot. Some people say that data can never tell us about causes. (I'm looking at you, David Hume). Most statisticians would say that certain specially designed experiments can tell us about certain kinds of causes, but of course there is much disagreement about which kinds of experiments and causes fit the bill.

I come to statistics through AI, so my position is the opposite extreme from Hume, which Gelman ably summarizes as: "a computer should be able to discern causal relationships from observational data, based on the reasonable argument that we, as humans, can do this ourselves in our everyday life with little recourse to experiment."

Still, this isn't a very popular view, so I thought it could use some advertising. A catchy jingle, maybe?

The birds do it,
The bees do it,
Even the rich folks on 5th avenue do it,
Let's do it,
Let's infer causes from purely observational data!

Comments are open, but this blog does sadly have a new comments policy:  all insults and ad hominem attacks will be deleted.  Keep it classy, internets.
Monday, February 1st, 2010
6:16 pm
* In watermelon car the deed was done and done again as my life is done in watermelon car.

* This is a sentence recommending that people who like self-describing works will enjoy This is the title of a typical incendiary blog post.

* John Cleese explains the brain.

* From the Canadian economics blog with the hilarious name of "Worthwhile Canadian Initiative", a supply & demand analysis of the Efficient Market Hypothesis.

* Interesting article by Kasparov on human and computer chess.
Saturday, January 23rd, 2010
4:33 pm
Book Reviews
While in San Diego a week ago, I had some time to read, which was lovely.

* The Suspicions of Mr. Whicher by Kate Summerscale. In 1860, one of the first detectives investigated a brutal murder in a Victorian England country home. This book describes both the twists and turns of the real life case, and how it essentially inspired the whole murder mystery genre. I enjoyed this book, but I was frustrated at times when Ms. Summerscale would interrupt the detectiving details for pages of social history or descriptions of how Dickens, Sir Arthur Conan Doyle, Wilkie Collins, or whoever stole this or that aspect of the situation for their work. I would also have preferred the brutal murder to not have been of a three year old boy, but I suppose I can't really blame the author for that.

* The Commitment: love, sex, marriage, and my family by Dan Savage. I've long been a fan of Dan Savage's Savage Love advice column (warning: NSFW), so I was happy to snatch this up for $1 from the Kensington Public Library Book Sale. This book is mostly a memoir about Savage's family and his deliberations with his boyfriend about whether to get married or not, with a middle section containing an extended argument for gay marriage. I loved the memoir parts, which were honest, insightful, and entertainingly written -- they made me really want to go out and get Savage's previous book about adopting his son (The Kid: What Happened After My Boyfriend and I Decided to Go Get Pregnant). The argument parts weren't bad, but neither were they original (Savage leans heavily on Andrew Sullivan's defense) and really there is only so much one can say when the other side has no coherent arguments of their own.

* The Jennifer Morgue by Charles Stross. Stross's The Atrocity Archives introduced the character of Bob Howard, a field agent for an NSA type organization set in a world where advanced mathematics and computer science can be used to summon demonic entities from other dimensions. (Because, after all, only such a situation could possibly justify the real life paranoia and secrecy that such organizations exhibit...) The Jennifer Morgue continues Bob's adventures and started off just fine, but soon devolved into an over-extended James Bond parody/tribute/memetic reflection thing. Still, it was enough of a page turner to get me through quite a few hours of flying, for which I am grateful.
Friday, January 8th, 2010
5:50 pm
Friday, December 18th, 2009
4:27 pm
Tis the time and/or season for linking
* It blows my mind, but George Lucas apparently wanted David Lynch to direct Return of the Jedi.

* "What's black and white and read all over?" As of this past December 13th, the Panda genome.

* How to Destroy The Earth. Full of delightful science facts such as this: when the Milky Way and Andromeda galaxies collide in 3 billion years, the expected number of stellar collisions is 6.

* Ross McKitrick came up with a great idea to resolve the climate change debate: tie carbon taxes to the troposphere's temperature. My posterior for expected climate change, btw, is about the same as Phil's.

* Awesome rube goldeberg machine (hat tip to David Miller).

* Libertarians for the French health care system.

* The always thought provoking Robin Hanson on married sex.

* I found Atul Gawande's New Yorker article on testing in agriculture extremely misleading. Specifically, he paints a picture in which the U.S.D.A. was a direct cause of increased U.S. agricultural productivity in the early 20th century. In reality, the exponential growth in agricultural productivity was an international phenomenon (it started in the U.K., in fact), started in the 19th century (or earlier), and was mainly driven by technologies (mechanization, four-field crop rotation, selective breeding, etc.) and social practices like enclosure.

* Finally, Shooting mosquitoes with frickin' lasers!
Thursday, December 3rd, 2009
1:57 pm
Simpson's Paradox
Hans and Gretel were two perfectly normal and happy children, except that their parents were both statisticians. This led them to having certain unusual experiences with higher probability than other children who were not so blessed.

For example, one day Hans and Gretel were given some paint to play with. Hans got a thimble of white paint from his mother, and a very large bucket of pink from his father, a pink which was perfectly balanced between red and white. Gretel got a thimble of red paint from her father, and from her mother she got a very large bucket of paint that was almost entirely white, with only a single drop of red mixed in.

Hans and Gretel like tables, so they made one:

From Momthimble of whitebucket of 99% white, 1% red
From Dadbucket of pinkthimble of red

Hans and Gretel were about to mix their respective paints together when O.J. Simpson jumped out from behind a bush. "Eeek!" screamed the children.

"Don't worry", Mr. Simpson replied, "I'm just here to make a prediction; namely that Gretel's paints when mixed together will be redder than Hans's. For you see, your Mom gave redder paint to Gretel, and so did your Dad."

"Eeek!" screamed the children again, both because they were still scared of O.J. Simpson and because their parents had taught them to fear erroneous conclusions. But since their parents had also taught them the power of experimental verification, they mixed their respective paints together and lo and behold: Hans had a bucket of pink while Gretel had a bucket of only slightly pinkish white.

O.J. was so chastened by this empirical disproof that he immediately turned himself into the authorities and confessed all of his crimes. Hans and Gretel had many other adventures, but never once were they tempted into aggregating subpopulation statistics as if they were masses.
Monday, November 9th, 2009
10:10 am
80% setup, 20% payoff

If Gertrude Stein were a gangster rapper and wrote a rhyming dictionary in which entries were given an ordering other than alphabetical:

nose before toes before bros before hoes before oreos before crows before a rose before a rose before a rose
[ << Previous 20 ]
My Website   About LiveJournal.com