Monday, February 27, 2012

Review: The City & The City

Unlike the other two China Mieville novels reviewed here, Perdido Street Station and The Scar, The City and The City is not set in the world of Bas-Lag. Instead, it's set in a contemporary world, somewhere in Europe. The story is about Inspector Borlu, who's assigned to investigate the murder of a young woman found in a park.

While ostensibly a detective novel, the novel is really about two cities, named Beszel and Ul Quoma. What's special about the cities are that they're super-imposed upon each other. Now, coming from his previous novels, I would expect there to be some fantastical explanation behind the super-imposition, but instead, half the mystery is figuring out the details of the super-imposition and how the two cities work, as well as the forces behind Breach, a power that operates to stop people from taking advantage of the super-imposition.

OK, so far so weird, which is good: Mieville is great at coming up with weird situations and then explaining all the details behind them. He works through all the implications of his own rules, involving the special training the cities have to give to visiting tourists so they do not accidentally Breach.

But then, the mystery gets into earnest and we get drawn behind the scenes to what's going on, and everything breaks down. Why? First of all, there's no reveal behind the nature of the super-imposition. We don't find out the history behind the two cities and their special relationship, nor do we ever see how it came to be. Furthermore, when the reveal shows up behind the nature of Breach and its enforcement, I at least, don't see how this could actually be a stable set up. It's quite clear to me that one city would have dominated the other through the course of human history, and we would have just one city and no Breach at the end. Even the resolution of the mystery makes no sense: the gains as depicted by the plot in the milieu could not possibly provide motivation for the characters involved!

If this was a first novel by an unknown author I might have been willing to brush all these problems away and say: "Great effort. Look for more stuff by him." But this is Mieville, and I feel cheated, as though he worked through all the mechanical parts of his plot device and setting, but didn't think through the implications of how historical forces would have acted to demolish this extremely unstable setup. While you might argue that "it's fantasy", I feel that the rigorous nature of the rules he's imposed on the setting as well as the nature of a mystery novel dictates that such logical inconsistencies not be overlooked.

I don't see how I could recommend this novel to anyone other than a die-hard Mieville fan, and of course, if you're one, you would have read it already.

Monday, February 20, 2012

Review: Liars and Outliers

My opinions about security are well known. Yet once in a while someone will point me to something by Bruce Schneier and I will read it and it would actually seem reasonable. So when I saw he had a new book out I read the sample and found myself ordering Liars and Outliers.

The trick? It's not actually a book about security. It's more or less a book about the Prisoner's Dilemma. Not the game that's played as a two player game, but in all its glories, with respect to an individual to society, and corporations against society, as well as all the complexity inside those edge cases.

I don't know how Schneier is as an engineer, but as a writer, he's far more interesting than any security person I've actually had to interact with as a co-worker. After reading this book, a lot of it I'm sure is that as an author, he's more likely to be willing to concede that many security considerations are less important than what's necessary to keep society functioning smoothly. Security engineers, on the other hand, often have to justify their jobs, and so you'll never hear security engineers say something like: "You're already too secure!" (And yes, it comes up --- as Steve Yegge points out in his famous blog posts, if you dial security down to zero, you get the Playstation Network, which is still somewhat useful, whereas if you dial security up to infinity, nobody uses it, and it's useless)

What's more interesting is that he says things like:
There is considerable evidence, both observational and experimental, that the group dynamics of a hierarchical organizational structure, especially a corporate one, dampen moral considerations as well. There are many reasons for this, and it seems to increase as organizations grow in size. (Pg. 169)
It's only a bit over the top to call corporations “immortal sociopaths,” as attorney and writer Joel Baken did. For corporations, the closest thing they have to morals is law. (Pg. 216)
What's interesting isn't just those quotes, it's that Schneier proceeds to explain why corporations, especially when they leave the startup stage, essentially turn evil and become sociopaths. What's really funny to me is that he uses Google frequently as an example of a non-evil corporation, especially the motto "Don't be evil", which was never actually a former corporate statement. Of course, the book went to press before the recent Google debacles became widely known. I'm not actually referring to the so-called privacy scandals, but to the fake pharmacy charges, where the federal government actually had a sting operation that showed that the policy that led to breaking the law went all the way to the top (including Larry Page), where not only did the executives knew they were breaking the law, they were explicitly told by the sting operators that they were breaking the law but approved and assisted anyway! It also explains why people who might otherwise be good human beings do regularly turn into sociopaths when employed by large corporations with lots of money. I gained a lot of sympathy for John T Reed's views as expressed in Succeeding as a result.

Schneier points out that 100% social conformance is not a good thing:
Increasing societal pressure isn't always worth it. It's not just the problem of diminishing returns discussed in Chapter 10. Looking back through history, the societies that enforce cooperation and conformance to the group norm, that ruthlessly clamp down and punish defectors, and that monitor every aspect of their citizens' lives are not societies we think of as free. (Pg. 245)

All in all, I read the book in just one night and found it fascinating and worth the time. Your views about society, cooperation, and how people behave (and misbehave) will change as a result of reading the book.


Friday, February 17, 2012

Review: Reamde

Reamde is the first Neal Stephenson book I've read since Cryptonomicon. A lot of this is because he seems to have forsworn all use of editors, which have made his books huge and ungainly without making them better.

The Hardback edition of Reamde runs 1056 pages, which is enough to kill someone if it was dropped off a tall building. Fortunately, I obtained the Kindle edition. The story revolves around a virus that encrypts all of a user's data on his hard drive with strong encryption, ransomed by dropping gold in a virtual world called T'Rain. The book goes into all the gory details about an MMORPG designed to make gold farming easy and respectable, leading you to believe that there might be some sort of plot involving virtual worlds.

Soon enough, though, Stephenson drops in Russian mafia, jihadhists, MI6 spies, survivalists, Hungarian hackers, and Chinese women, in rapid succession and in higly improbable situations, leading the reader to sort out where the plot is going. The plot then gyrates off one improbable situation after another, leading the characters to diverge and then converge finally, all in one big battle with lots of bloods, guts, and loving description of military hardware.

We do get a happy ending, and everything's tied off nicely in a bow. But the ungainliness of the plot and the improbability all makes you wonder how if this was mainstream fiction, how anybody could consider science fiction or fantasy "speculative". Altered Carbon feels like hard-boiled realistic fiction compared to this stuff.

I spent about 10 days working through this book, and I'm not sure I got very much out of it. It's brain candy, and the feeling I got after reading this book is the feeling I'd get if I were to down an entire shipping container full of hostess twinkie. A lot of artificial ingredients that's ultimately not very nutritious or satisfying.

Not recommended.

Thursday, February 16, 2012

Review: Logitech Wireless Keyboard K320

My 6 year old Kinesis keyboard started getting flakey: whenever I first started up my machine, it wouldn't be recognized 50% of the time, requiring me to unplug and plug in the USB plug. I considered replacing it with another Kinesis, since I liked the keyboard, but two things stopped me: first the cost ($300 is a lot of money for a keyboard), and secondly, my wife complained that my keyboard was quite noisy. Since I was using the Kinesis as a matter of preference, rather than because I was having trouble with repetitive stress injuries, I thought I would try something cheaper.

First, I bought a Logitech K250 at an incredibly cheap price ($10) from one of the daily deals site. I was disappointed with the keyboard feel: the keys feel squishy. My wife, however, liked it so she grabbed it for use with her laptop.

I briefly considered the Logitech K750, but the chiclet style keys had me terrified. I've never enjoyed chiclet style keyboards: I detest Apple laptop keyboards, for instance. The lack of travel on the keys never give me sufficient positive feedback, so I end up with far more typos than I normally have.

Finally, when the Logitech Outlet store had a special on the K320, I decided that I liked the look of it enough to try one for $20.

The box comes with a usb unifying adapter, which is tiny, and a 90 degree adapter in case your USB slot would be blocked by the unifying adapter. I'm disappointed that there are no cheap bluetooth keyboards that are made for anything other than an ipad, since that would work out best, but the nice thing about these proprietary dongles is that they work even when the PC isn't booted into the OS, for tweaking with the BIOS, etc.

The keyboard feel is excellent, a little reminiscent of the old IBM PC keyboards, with a full size travel and a decent stroke and positive and firm feedback. The negative is that the keyboard's wider than my old Kinesis, mostly because of the numeric keypad. It's also quite broad, with a bunch of buttons that I might never use. I used to dislike the Windows key on keyboards, but now that I have a multi-monitor setup, it's frequently useful for a few shortcuts.

All in all, I've retired my Kinesis for quite a few days now, and so far it seems quite serviceable. At the very least, it's not $300. Recommended.

Sunday, February 12, 2012

Review: Grand Pursuit - The Story of Economic Genius

I'll admit it. I'm an economics junkie, and enjoy even reading books like Minsky's Stabilizing an Unstable Economy. So I was excited when Sylvia Nasar of "A Beautiful Mind" tackled the history of great economic minds in Grand Pursuit.

By far the biggest disappointment is that with only limited space and spanning this much time, Nasar could only grant superficial coverage of many of the ideas. I was even more annoyed when she insisted on wasting precious pages on boring genealogy details rather than the big ideas of the man.

That said, Nasar does a passable job of describing economic history. Her description of Karl Marx as a person who'd never bothered to learn English despite writing his book there, and who never even visited a single factory to do his work. Several not very well known economists are covered, including Alfred Marshall, Beatrice Webb, and Joan Robinson. Strangely enough, she never talks about Adam Smith except in passing.

By the time we get to the greats like Schumpeter, Keynes, Friedman, and Samuelson, the repeated mini-biographies are starting to wear thin. However, this is where the action starts, so I was quite perked up. I learned quite a bit more about Keynes that I didn't know before. Nasar also "gets" Keynes, though she doesn't quite take the pains to explain why Keynes' General Theory was widely misunderstood, even more so than say, Einstein's.

The war years are covered in great detail, though big breakthroughs (like Irving Fisher's realization of the relationship between interest rate, the business cycle, and inflation) don't quite get the headline attention they deserve: sometimes you feel as though Nasar is more interested in Keynes' bi-sexuality than in his ideas.

What does come across to me as brand new information is the section on Hayek. For instance, he and Keynes were friends and supported each other's work. Hayek is definitely not the libertarian that his later followers make him out to be, and Nasar delights in pointing that out, especially when Republicans tried to court him and get his approval.

Ultimately, while the book was worth reading for me, I wonder how many non-economics junkies will be able to keep their eyes open during the long diversions. Not really recommended.

Wednesday, February 08, 2012

Ethical Investing

When you talk about ethical investing, people naturally conflate it with socially responsible investing.

The thing is, nowadays, the financial industry is part of the problem. As Lawrence Lessig described in Republic, Lost, part of the reason carried interest, for instance, is taxed as capital gains instead of income is because of all the lobbying the financial industry did. And who can forget the financial industry's successful lobbying of the repeal of Glass-Steagall, which indirectly led to the financial crisis of 2008.

I'm a firm advocate of not having a financial adviser, but many people choose to use them. (I think it's pretty silly given how little time I actually devote to my finances) If you do choose to use one from the big white shoe firms like Morgan Stanley, Merrill Lynch, or Fidelity, you're in turn contributing to their lobbying efforts and their continual attempts to undermine the "main street economy."

If you must use a financial adviser, I recommend Vanguard, which is a non-profit, or Wealthfront, which is a Silicon Valley startup that currently doesn't have the cash to do any lobbying.

Now, lots of people brag to me about how much hand-holding they get from their white shoe adviser. What's interesting to me is that they're all implying that the service I get from Vanguard must be sub-standard and do-it-yourself. They couldn't be further from the truth. I'll illustrate with an example: several years ago I took a foreign assignment in Munich. While I was there, another visitor from Mountain View showed up on assignment as well. I asked if he needed any help dealing with the banking system in Munich, and he said, "Oh no. Bank of America says they're affiliated with Deutsche Bank, and everything should go smoothly." 2 weeks later he was in the office panicking: he had 3 days to provide a security deposit to his prospective landlord to get his apartment, and nothing had happened. I told him to call Vanguard. Within 2 hours, all his problems were solved, and his security deposit was ready. Now, if this story was about me, you might say, "Sure. Piaw's a Flagship customer, so he gets special attention." But this was for someone who didn't qualify for Flagship. Nevertheless, Vanguard moved heaven and earth to solve his problems, getting his money wired overseas for no fee whatsoever.

Of course, there're still the folks who brag about their financial advisers sending them ice-cream. Given the price difference between what they're paying and what they could pay Vanguard instead (Vanguard's financial advise fee is $250/year, fixed) or even Wealthfront, I calculate that to be $10,000/gallon worth of ice cream. No wonder it tasted so good!

The financial industry is unlike any other industry on the planet: it's the only industry in which the less you pay, the more you get. If you don't already have effectively infinite money, it pays to pay attention to that little detail.

Monday, February 06, 2012

A post-IPO sale simulator

A few years ago, I wrote a long blog entry on how to appropriately diversify away from company stock. While the article addressed an employee working at a large, post public company, nearly every employee that works at a pre-IPO startup (such as the social media startups that have gone public recently or will be going public soon) has to consider what their diversification strategy is.

In general, how worried you should be depends on how much money you've made. If you were early enough at Google, LinkedIn, or Netflix, you're in the "filthy rich" category, and you really don't care when you sell or how it happens because unless the company you work for is a WebVan, you have way more money than you can spend. (Note that some times it's very difficult to tell, since Ariba was a viable company but that didn't stop its employees from making some very bad decisions --- An Engineer's Guide to Silicon Valley Startups actually had a few case studies of phenomenally poor decision making)

However, if you're on the edge of serious money (e.g., enough money to never have to work again --- which is far less money than most people assume), then figuring out when to sell is really important. You might think that as an insider for however many years, you ought to have better insight than the market. To some extent, you could be right: when I saw that Mpath/HearMe went public, I knew that the company was not profitable. But that didn't stop the dot com bubble from inflating the stock to well past $50/share before crashing. By the way, one reason to not aim for the moon as far as net-worth is concern is that as John T Reed points out in his book Succeeding, doing so distorts your risk-reward perspective and causes you to take un-needed risk, whereas a more prudent and safer strategy could have netted you more money sooner. Yes, I learnt that the hard way during the dot-com bubble.

I knew that Google was extremely profitable, but because its IPO auction process pissed-off Wall Street, the IPO started out vastly under-valued. Yet, when it came time to really sell Google stock, I found it to be one of the hardest decisions I'd ever made in my life. (One interesting thing about writing that blog post was that many Google employees mis-understood me and thought that I didn't sell any at $700 --- but I was happy to let them think that)

Wealthfront today has launched an IPO sale simulator. I was provided a pre-release version of it and played with it quite a bit, and it does a great job showing you all the different what-ifs scenarios and implications of different strategies for different companies. To simplify things, Wealthfront only has 4 different strategies: sell 10% each quarter for 10 quarters, sell 10% of remaining shares each quarter, sell 50% up front and 10% there after, and sell all after the lockup. My own strategy, which is: "sell something every year but let the price determine how much you sell" obviously can't be represented in any of the mechanical strategies.

One of the things the wealthfront blogpost covers is that if you wait for the second month of a quarter to sell, there's quite a bit less volatility. I think that doesn't quite work: of my 3 IPOs, Google and Pure had very tight trading windows which prevent employees from utilizing such strategies without special effort. In particular, it was possible for me to make far more money from Google stock when not being an employee because I was now free to trade!

Finally, the toughest part about post-IPO selling is that you don't know apriori whether your stock's going to do what Google did, or what DivX or Netflix did. And yes, I do own pre-IPO stock in Facebook, and I have a selling strategy in mind, but I'm not telling, at least not on this blog.

Friday, February 03, 2012

A Surprising Change in Google+ Engagement

In December, I wrote a piece about how Google+ engagement was surprisingly low, considering how many people followed me on Google+ versus Facebook and Quora.

Well, I went back to look at the month of January and wow, what a change a couple of months have made. Google+ is now right on top of my referrals at 262 visits, versus Facebook at 235. Quora is in 3rd place at 168, followed by linkedin. Of course, Google's organic search trumps everyone at 3000 visitors over the same period.

I have no idea what's caused the change, though there is one clue: new visitors from Google+ comprise a much smaller percentage of the referred volume than they do from Facebook and Quora. What this means is that most of my friends (a lot of Google affiliated people) have migrated over to Google Plus, probably from Google Reader, since Reader no longer has any social features.

I'm not sure what this means in the long haul: I suppose I could duplicate-post my current Delicious Feed onto Google Plus for a bit to see if engagements goes up even further, but my suspicion is that it will have zero impact.

Regardless, it's clear: with engagement going up over the last 2 months, I cannot ignore Google+, even though I would have much preferred Friendfeed to win, for instance.

Thursday, February 02, 2012

Review: Buzz, The Science and Lore of Alcohol and Caffeine

The two most common drugs in use in the world are alcohol and caffeine. Buzz is a book that tries to explain how both substances work, how they affect your brain and body, and dispels some misconceptions as well as re-affirming some wisdom about what happens.

The book is short (which is a feature) and easily read in an afternoon. The section on alcohol is interesting, explaining in detail how potent a drug it is: fundamentally, the amount of alcohol consumed in even one drink is enough to push it around in your blood stream. It also explains why you get drunk less easily on a full stomach, as well as the complex effects alcohol has on your brain, most of which isn't actually any good. He also discusses the effects of red wine, along with other beneficial health benefits of regular red wine consumption, though again with the caveat that if you don't already drink, you shouldn't start drinking in order to get the health benefits. There's also an extensive analysis of why some people drink, and how natural variances in preferences lead to some people who strongly prefer alcohol and some people don't.

Caffeine to me is the more interesting of the two drugs in question. Unlike alcohol, it's not strongly related to any sociological problems, but can also be addicting, as many of my friends who've gotten used to its effects and then tried to withdraw cold-turkey can attest. What's amazing to me is how long the half life of caffeine is in the body. It's 6-7 hours in adults, and twice that long in infants and small children. What's also amazing is that there's huge variance in people, so for some people drinking coffee in the morning can lead to a sleepless night. There's also an explanation of how caffeine works in your body and a short history of caffeine doping in the world of sports. There's an interesting section on how caffeine affects PMS, as well as another section on the effects of caffeine on fetuses. Of more practical use, there's an explanation of what kind of tasks caffeine helps with, and what caffeine fails to do. The drug is in fact complex and interacts in a complicated way with human bodies. (An interesting section in the book explains why caffeine is found in so many different plants --- it is essentially a self-defense mechanism for the plant!)

All in all, the book's well worth your time. It's well written, though not brilliantly so. Recommended.