Earlier this month, the Onion A.V. Club ran a blog post about pop-culture deal-breakers...that is, what music/movies/books/etc. would cause you to immediately break up with your significant other. I love reading through the comments of these kind of posts because it gives me a feel for what the masses like and think. Well, maybe just the internet-saavy hipster douchebags. Anyway, it seems like the majority say that not liking Prince is a deal-breaker. I don't really like Prince, but I can still appreciate him and I'll listen if somebody else wishes to. I thought his recent Super Bowl performance was pretty good. A lot of people also mentioned The Simpsons because that show represents a certain sense of humor that people identify with. I can't think of a single person in my generation that can't at least appreciate the Simpsons on some level.
Anyway, I'm not really a deal-breaker kind of guy, but I've been thinking about this off and on, and I've come up with two.
1. Reading horoscopes. Period.
2. Dave Matthews Band
I know people who like both of these, but I would have a very difficult time dating somebody who likes them. Even reading horoscopes "for fun" gives some credence to that bullshit, and if you truly don't believe in them, they are boring as hell and a waste of time. As for DMB, aside from the utterly crappy music, lyrics, and singing, their fans are mostly annoying as well. I think my idea of hell is the first 10 seconds of Ants Marching on repeat. Apologies to my old office-mate Joe, who, aside from liking DMB, is a really cool guy.
Friday, October 24, 2008
Monday, October 20, 2008
Black Holes
So I finally finished Black Holes and Time Warps the other day. I've had it for a very long time--it was one of the first books I got as part of the 5-for-$1 deal when signing up for the book of the month club way back in high school (once I began earning disposable income). I almost got rid of it with a bunch of other books when moving out to Virginia, but I decided to actually read it first before I try to sell it back. It wasn't really worth the wait.
One of my biggest peeves with the popular science non-fiction genre is how dumbed down everything is. And all the books about relativity and cosmology use the same old analogies that break down when you're talking about actual real-world situations. Kip Thorne's book is different in that he tries to explain the nitty gritty details about Einstein's equations, relativity, and black holes, and he uses the actual analogies that physicists use when talking to each other, not the dumbed down analogies. However, Thorne still utilizes my second peeve: the lack of equations. For people like me, sometimes a simple equation makes a concept crystal clear. About halfway through the book, Thorne runs out of straightforward explanations and analogies and is forced to start making blanket claims about black holes, and at that point I start asking "why?". How does topology prove every black hole has a singularity? How can a spinning black hole create a "naked singularity" (a black hole that is not "black")? Reading Thorne's labored descriptions of how black holes must only "grow larger" or how they "radiate" make me wish he could include a few straightforward equations to show why this is so. The whole chapter about gravitational waves without a single equation made me cringe. Some equations are found in the footnotes, and it's funny to read a page and a half of explanation and then to read a two sentence footnote with an equation that makes everything finally clear.
There are three things I liked about the book. First, it finally convinced me that black holes are real and not simply a mathematical toy that physicists like to play with. Well, the recent Nova episode helped, where they actually observed stars getting flung around a blank spot in the center of our galaxy. Secondly, the book placed everything in a historical context. It could almost be considered a history of science book because it credits everybody for their individual contributions to the field. It also shows how the field built upon the ideas of previous researchers. That was pretty interesting, and I finally was able to hook some names with ideas, and to also learn new names which I'm noticing all the time now (e.g. Jansky Fellowship for radio astronomy). And third, Thorne makes it explicit (albeit later in the book) that the hyperspace dimension in which spacetime is bent by mass is NOT real, but only a mathematical construct. I really hate it (especially in quantum mechanics) when authors take equations and interpret them literally when that is not their purpose. Thorne even takes the time to explain that in certain situations, physicists use a spacetime-warping model for some situations and a "flat" spacetime-compressing model for other situations, just as how we treat subatomic particles as particles or waves as the situation requires. I applaud him for that.
So bottome line, I would say the first ~300 pages or so (up to chapter 8) of the book are a pretty good thorough popular science introduction to relativity and black holes. I wouldn't necessarily say it's an exciting read (chapter seven was kind of boring, even though it had a lot of cool concepts in it), but it explains black holes pretty well. There are two nagging questions I still have. First, I would like a better explanation of how the black holes grow and shrink (can black holes disappear if nothing is added to them?). Second, the book does a pretty good job explaining why only stars of sufficient mass can become black holes (smaller stars become neutron stars for white dwarfs) but I keep hearing things about tiny black holes, especially the kind that can be created once the LHC is finally operational. So how can a black hole exist below the Tolman-Oppenheimer-Volkoff limit?
Slightly off topic, I am still waiting for that popular science book that is geared for people with a good algebra background. Most books are either way too dumbed down (with no equations) for actual textbooks (= dry and boring with too many equations and not enough explanation). Maybe something with normal text for the layman and side boxes full of equations for the scientist. Actually Shadows of the Mind by Roger Penrose did a really good job of that. But then Penrose starts talking about M-branes and I get lost again. Maybe I should try reading that book again, or maybe his most recent monstrosity.
One of my biggest peeves with the popular science non-fiction genre is how dumbed down everything is. And all the books about relativity and cosmology use the same old analogies that break down when you're talking about actual real-world situations. Kip Thorne's book is different in that he tries to explain the nitty gritty details about Einstein's equations, relativity, and black holes, and he uses the actual analogies that physicists use when talking to each other, not the dumbed down analogies. However, Thorne still utilizes my second peeve: the lack of equations. For people like me, sometimes a simple equation makes a concept crystal clear. About halfway through the book, Thorne runs out of straightforward explanations and analogies and is forced to start making blanket claims about black holes, and at that point I start asking "why?". How does topology prove every black hole has a singularity? How can a spinning black hole create a "naked singularity" (a black hole that is not "black")? Reading Thorne's labored descriptions of how black holes must only "grow larger" or how they "radiate" make me wish he could include a few straightforward equations to show why this is so. The whole chapter about gravitational waves without a single equation made me cringe. Some equations are found in the footnotes, and it's funny to read a page and a half of explanation and then to read a two sentence footnote with an equation that makes everything finally clear.
There are three things I liked about the book. First, it finally convinced me that black holes are real and not simply a mathematical toy that physicists like to play with. Well, the recent Nova episode helped, where they actually observed stars getting flung around a blank spot in the center of our galaxy. Secondly, the book placed everything in a historical context. It could almost be considered a history of science book because it credits everybody for their individual contributions to the field. It also shows how the field built upon the ideas of previous researchers. That was pretty interesting, and I finally was able to hook some names with ideas, and to also learn new names which I'm noticing all the time now (e.g. Jansky Fellowship for radio astronomy). And third, Thorne makes it explicit (albeit later in the book) that the hyperspace dimension in which spacetime is bent by mass is NOT real, but only a mathematical construct. I really hate it (especially in quantum mechanics) when authors take equations and interpret them literally when that is not their purpose. Thorne even takes the time to explain that in certain situations, physicists use a spacetime-warping model for some situations and a "flat" spacetime-compressing model for other situations, just as how we treat subatomic particles as particles or waves as the situation requires. I applaud him for that.
So bottome line, I would say the first ~300 pages or so (up to chapter 8) of the book are a pretty good thorough popular science introduction to relativity and black holes. I wouldn't necessarily say it's an exciting read (chapter seven was kind of boring, even though it had a lot of cool concepts in it), but it explains black holes pretty well. There are two nagging questions I still have. First, I would like a better explanation of how the black holes grow and shrink (can black holes disappear if nothing is added to them?). Second, the book does a pretty good job explaining why only stars of sufficient mass can become black holes (smaller stars become neutron stars for white dwarfs) but I keep hearing things about tiny black holes, especially the kind that can be created once the LHC is finally operational. So how can a black hole exist below the Tolman-Oppenheimer-Volkoff limit?
Slightly off topic, I am still waiting for that popular science book that is geared for people with a good algebra background. Most books are either way too dumbed down (with no equations) for actual textbooks (= dry and boring with too many equations and not enough explanation). Maybe something with normal text for the layman and side boxes full of equations for the scientist. Actually Shadows of the Mind by Roger Penrose did a really good job of that. But then Penrose starts talking about M-branes and I get lost again. Maybe I should try reading that book again, or maybe his most recent monstrosity.
Thursday, October 09, 2008
Behind The Bailout
All this recent shit with the market is really affecting my libertarian values about the economy and markets. It's been gradually shifting over the years, but the last nail has been pounded into the coffin of "markets regulate themselves". I first really learned about Adam Smith and the "invisible hand" in my high school AP European History class, and at the time I really did believe that, if left alone, markets would establish a fair and stable system, although I didn't believe that anything (like God) actively regulated it. I think now we have a good case study disproving that idea. However, I still don't necessarily disapprove of deregulation (try untangling those negatives!). What I want is an "open" economy, not in terms of free international trade, but in terms of complete disclosure (which is a loose form of regulation, I guess). Every day in the news I hear that banks aren't lending any money, and today I found out why. On the long drive up to the synchrotron, I listened to the podcast of this week's This American Life, which I really strongly advise everybody to take a listen because it does a pretty decent job of explaining what went wrong and why we need a bailout. Basically the problem is with the credit default swap market. A CDS is when I have a bond from, say, Lehman Brothers, which is basically a loan to the company, or my investment into the company. I don't like that Lehman Brothers is making risky investments into subprime mortgages, and I want to make sure I don't lose all my money if they tank, so I go to a bank or a hedge fund or some other wall street company and make a contract to pay 2% of the bond value every year and in return, if something happens to Lehman Brothers, I get my initial investment. So I'm paying $2k a year to make sure I can at least walk away with my initial $100k. Most wall street entities do this, and I see no problem with it. But I don't even have to have made an initial investment with Lehman Brothers! I can just go to AIG and pay them $2k a year for a CDS in the hopes that Lehman Brothers will tank and then I'll get $100k. Let's say 100 people all decide to do this. AIG is now raking in the dough, $200k a year, and the chances of Lehman Brothers failing is small. But Lehman Brothers did fail, and now they have to pay out $10 million, which is why the government had to bail them out. The CDS market is completely deregulated. There is no requirement to actually hold the bond you are buying the CDS for, and the estimate I heard today is that the value of the CDS market is ten times the total real value of the bonds in the contracts. When people hear this they want to regulate in two ways:
1. In order to buy a CDS, you must hold the bond in question. In other words, a CDS is solely an insurance policy and not a bet that a company will fail.
2. In order to sell a CDS, you must be able to pay the full amount in case of default. In other words, AIG has to stop selling CDSs when they run out of capital to pay out for them.
I don't think we need to go regulate CDSs in this way. Instead:
3. Buying and selling of CDSs will occur in an established market with full disclosure
You see, in the above example, when Lehman Brothers files for bankruptcy, all the CDSs have to be payed, and AIG goes under. End of story. This does not affect the economy as a whole, except for AIG customers. But there is a CDS network. All those people who bought a 2% Lehman Brothers CDS from AIG realize that simply paying $2k a year on a bet is not a smart investment, so they've resold their CDS to somebody else for 2.5% or 3% (as Lehman Brothers gets shakier and more likely to fail). So those people pay $2k to AIG, but get $3k from somebody else. If Lehman Brothers fails, they owe the third party $1ook, but they will get that from AIG. But this third party has sold the CDS to a fourth party, and so on. There is this giant chain or network of CDS buying and selling. And if AIG fails because they have to pay out so much, AIG can't pay the $100k, and nobody gets paid. Also, other people have bought and sold CDSs against AIG, so the failure of AIG would cause another massive payout. Remember, the total value of CDSs sold is 10 times the total real value of the bonds involved, so this would cause a systemic failure of wall street.
Here's why I think solution #3 is the best solution. Because the CDS market is undisclosed, nobody knows what companies or fund managers are buying and selling. Nobody knows how much CDSs AIG has sold, meaning nobody knows how much AIG is fucked if Lehman Brothers goes down. If they knew, they wouldn't buy another CDS from AIG (although they probably would buy a CDS against AIG). Also, we have no idea how interconnected the CDS network is. But if the CDS market was fully disclosed and public, then investors would have a much better idea which CDS investments are better than others.
This takes us back to Adam Smith. Markets will "self regulate" only when both the buyer and seller have equal knowledge. Let's say I want to buy a mineral from a dealer at a gem and mineral show. As a geologist, I know how rare certain minerals are, and both the dealer and I know how much the specimen is worth, and we bargain until we reach a fair price. If I don't know anything about geology or mineralogy, the dealer may exaggerate how rare this particular mineral is, and I may overpay. If the dealer is just a hippie into crystals and that kind of bullshit, I may be able to talk down the value of the mineral (or recognize it as something else altogether, which has happened to me). So supply and demand are not the only factors that determine price. Information and knowledge are very important. Let the American people and investors know how much potential debt a company is getting themselves into and how interconnected they are to others, and this will have a big effect on the market.
So that's the justification for the AIG bailout. Soon the CDS market will be regulated to some degree. However, what about the big $700 billion bailout just passed in Congress? That's more directly related to the housing crisis to keep more mortgage companies (and mortgage-backed securities companies) from going under, although it is affected somewhat by the CDS crunch and the commercial paper freeze. I'm still not sure what my feelings are on this one. On the one hand, I think it's stupid for the government to buy mortgages and mortgage backed securities to take them off the books because we will have a hard time selling them off later (and there's a chance the American taxpayers can get really screwed on this deal). On the other hand, the bailout bill allows the Secretary of the Treasury to invest in banks instead of buying securities, taking partial ownership like the bailout of AIG. I like this idea better, but I'm wary of the government getting too involved in business, and there's potential for the loss of a lot of taxpayer money that way.
OK so I know this post was long, but the bottom line for me is that I'm fine with wall street making their money by clever schemes, risky ventures, and confusing investments, but all that trading information should be public because it can apparently affect everything else. Open markets and open government, that's what I say.
Also, I want to mention that market "self regulation" cannot occur if bailouts are allowed. The current CDS situation is being "self regulated" out of existence because it is untenable, but the bailouts will allow it to continue, therefore we are obligated to regulate it.
OK, that's enough!
1. In order to buy a CDS, you must hold the bond in question. In other words, a CDS is solely an insurance policy and not a bet that a company will fail.
2. In order to sell a CDS, you must be able to pay the full amount in case of default. In other words, AIG has to stop selling CDSs when they run out of capital to pay out for them.
I don't think we need to go regulate CDSs in this way. Instead:
3. Buying and selling of CDSs will occur in an established market with full disclosure
You see, in the above example, when Lehman Brothers files for bankruptcy, all the CDSs have to be payed, and AIG goes under. End of story. This does not affect the economy as a whole, except for AIG customers. But there is a CDS network. All those people who bought a 2% Lehman Brothers CDS from AIG realize that simply paying $2k a year on a bet is not a smart investment, so they've resold their CDS to somebody else for 2.5% or 3% (as Lehman Brothers gets shakier and more likely to fail). So those people pay $2k to AIG, but get $3k from somebody else. If Lehman Brothers fails, they owe the third party $1ook, but they will get that from AIG. But this third party has sold the CDS to a fourth party, and so on. There is this giant chain or network of CDS buying and selling. And if AIG fails because they have to pay out so much, AIG can't pay the $100k, and nobody gets paid. Also, other people have bought and sold CDSs against AIG, so the failure of AIG would cause another massive payout. Remember, the total value of CDSs sold is 10 times the total real value of the bonds involved, so this would cause a systemic failure of wall street.
Here's why I think solution #3 is the best solution. Because the CDS market is undisclosed, nobody knows what companies or fund managers are buying and selling. Nobody knows how much CDSs AIG has sold, meaning nobody knows how much AIG is fucked if Lehman Brothers goes down. If they knew, they wouldn't buy another CDS from AIG (although they probably would buy a CDS against AIG). Also, we have no idea how interconnected the CDS network is. But if the CDS market was fully disclosed and public, then investors would have a much better idea which CDS investments are better than others.
This takes us back to Adam Smith. Markets will "self regulate" only when both the buyer and seller have equal knowledge. Let's say I want to buy a mineral from a dealer at a gem and mineral show. As a geologist, I know how rare certain minerals are, and both the dealer and I know how much the specimen is worth, and we bargain until we reach a fair price. If I don't know anything about geology or mineralogy, the dealer may exaggerate how rare this particular mineral is, and I may overpay. If the dealer is just a hippie into crystals and that kind of bullshit, I may be able to talk down the value of the mineral (or recognize it as something else altogether, which has happened to me). So supply and demand are not the only factors that determine price. Information and knowledge are very important. Let the American people and investors know how much potential debt a company is getting themselves into and how interconnected they are to others, and this will have a big effect on the market.
So that's the justification for the AIG bailout. Soon the CDS market will be regulated to some degree. However, what about the big $700 billion bailout just passed in Congress? That's more directly related to the housing crisis to keep more mortgage companies (and mortgage-backed securities companies) from going under, although it is affected somewhat by the CDS crunch and the commercial paper freeze. I'm still not sure what my feelings are on this one. On the one hand, I think it's stupid for the government to buy mortgages and mortgage backed securities to take them off the books because we will have a hard time selling them off later (and there's a chance the American taxpayers can get really screwed on this deal). On the other hand, the bailout bill allows the Secretary of the Treasury to invest in banks instead of buying securities, taking partial ownership like the bailout of AIG. I like this idea better, but I'm wary of the government getting too involved in business, and there's potential for the loss of a lot of taxpayer money that way.
OK so I know this post was long, but the bottom line for me is that I'm fine with wall street making their money by clever schemes, risky ventures, and confusing investments, but all that trading information should be public because it can apparently affect everything else. Open markets and open government, that's what I say.
Also, I want to mention that market "self regulation" cannot occur if bailouts are allowed. The current CDS situation is being "self regulated" out of existence because it is untenable, but the bailouts will allow it to continue, therefore we are obligated to regulate it.
OK, that's enough!
Wednesday, October 01, 2008
Zotero
In my line of work, good reference management is essential. In grad school I had an elaborate system of manila file folders organized by subject, in which I had copies and printouts of every paper I might need to cite. Needless to say that system became cumbersome quickly, especially once I started to work in multiple fields of research.
At NRL EndNote was available to me, so I spent quite a bit of time entering all my references into a database. It was then I realized how useful reference software could be. Not only were my references all searchable now, I was able to automatically cite references and create a bibliography automatically in Word (using a plugin). These are the two criteria upon which to judge good reference software: 1) ease of database management (adding references, searching references, organizing references, etc.) and 2) ease of use in Word (citing references, building bibliographies, editing reference styles, etc.). EndNote is actually pretty great at #2. I recently submitted a paper to Science, and after changing style settings in the EndNote Software, I only had to make one edit to my bibliography list (instead of having to make a bunch of edits for almost every reference, which is the biggest complaint I hear about reference software). But EndNote leaves a bit to be desired in #1. It looks and acts clunky. I do a lot of searching for references online, and it's not always straightforward to import into EndNote. I'm always having to manually enter the DOI and manually save a PDF copy of each paper. Also, EndNote costs money, and I would like to take my reference database with me when I leave NRL, so I've been looking for cheap alternatives.
Enter Zotero. It's a firefox plugin developed at George Mason University, and it accomplishes #1 almost perfectly. When I find a paper I want to add online, all I have to do is click an icon and it downloads all the information (including DOI) and in most cases also a copy of the PDF. I'm already searching for references in my browser, so it makes perfect sense to manage my references right there in my browser. Zotero does organization and search much better than EndNote. It allows for tags as well as directories, so you can organize everything the way you want to. And Zotero allows you to search text in PDFs! Unfortunately Zotero is not as great with #2. It has a buggy Word plugin, and their citation style settings have errors in them that can't be edited. Also, EndNote did not export DOI numbers or all the PDFs I had downloaded, so I basically have to go through the entire database in Zotero and manually save in the PDFs. However, it turns out it's much quicker just to look up the paper again online and import it directly into Zotero than to edit the existing entry.
I'm toying with the idea of using Zotero for reference management, and then when I write I'll export to EndNote so I can use the great EndNote plugin in Word, but that sounds like way to much work for what it's worth.
At NRL EndNote was available to me, so I spent quite a bit of time entering all my references into a database. It was then I realized how useful reference software could be. Not only were my references all searchable now, I was able to automatically cite references and create a bibliography automatically in Word (using a plugin). These are the two criteria upon which to judge good reference software: 1) ease of database management (adding references, searching references, organizing references, etc.) and 2) ease of use in Word (citing references, building bibliographies, editing reference styles, etc.). EndNote is actually pretty great at #2. I recently submitted a paper to Science, and after changing style settings in the EndNote Software, I only had to make one edit to my bibliography list (instead of having to make a bunch of edits for almost every reference, which is the biggest complaint I hear about reference software). But EndNote leaves a bit to be desired in #1. It looks and acts clunky. I do a lot of searching for references online, and it's not always straightforward to import into EndNote. I'm always having to manually enter the DOI and manually save a PDF copy of each paper. Also, EndNote costs money, and I would like to take my reference database with me when I leave NRL, so I've been looking for cheap alternatives.
Enter Zotero. It's a firefox plugin developed at George Mason University, and it accomplishes #1 almost perfectly. When I find a paper I want to add online, all I have to do is click an icon and it downloads all the information (including DOI) and in most cases also a copy of the PDF. I'm already searching for references in my browser, so it makes perfect sense to manage my references right there in my browser. Zotero does organization and search much better than EndNote. It allows for tags as well as directories, so you can organize everything the way you want to. And Zotero allows you to search text in PDFs! Unfortunately Zotero is not as great with #2. It has a buggy Word plugin, and their citation style settings have errors in them that can't be edited. Also, EndNote did not export DOI numbers or all the PDFs I had downloaded, so I basically have to go through the entire database in Zotero and manually save in the PDFs. However, it turns out it's much quicker just to look up the paper again online and import it directly into Zotero than to edit the existing entry.
I'm toying with the idea of using Zotero for reference management, and then when I write I'll export to EndNote so I can use the great EndNote plugin in Word, but that sounds like way to much work for what it's worth.
Subscribe to:
Posts (Atom)
