Month long cycling blog..!

One of the nice perks in working for LSI is a one month paid sabbatical leave you get once in 10 years  to go do whatever you want. Some of my colleagues who had taken the leave have tried interesting things like a month long road trip in US. Now it is the turn of another colleague named Brad Grande, who has decided to spend the entire month bicycling 1500 miles along the Alcan Highway in Alaska..! He has been chronicling his adventure (starting from months back when he started preparing for it) at http://bradsalcanadventure.blogspot.com.

Check it out when you get a chance. 

I had done two smaller cycling trips as part of NCC during my ugrad years circa 1985. One trip was from Chidambaram to Bangalore and back. Another one was from Chidambaram to Tirupathi and back. Each one lasted about 7 days with an additional 3 days break in between (in Bangalore & Tirupathi). Naturally I never knew enough in those days to do practice runs during the previous days/weeks, get special bike designed for long distance bicycling, etc. So, we prepared just a basic luggage with change of clothes and toothbrush and started pedaling with a very ordinary cycle. No GPS or any pre-planning for boarding/lodging. When it got dark we will stop in the town/village we were in and eat in the local restaurant/t-shop/home that offered food and slept in schools, rice mills, etc. It was memorable. See if you can locate me in these pictures taken during the Bangalore trip. :-)

Brad's adventure is quite the opposite. He has a bike that costs some $3000. He has a guide and another friend biking with him with well thought out luggage, travel route, GPS, cell phone, laptop, internet access to blog everyday, etc. Pretty cool to see him spend the whole month doing this. I should be eligible for a sabbatical in 2014. Wonder what I should do..?!

Book Review: The Age of Turbulence

Finished reading Alan Greenspan's "The Age of Turbulence: Adventures in a New World" during the long (July 4th) weekend. Since I have been interested in macro economic policies, found these 500 page book to be quite an interesting read though I am as usual embarrassed to admit the the time it took to read it end to end. Greenspan has been in the field of economics for half a century, initially as a private consultant and then for several decades as a federal govt appointee. He has served as the Federal Reserve Chairman for more than 18 years spanning Ronald Reagan (who appointed him), George H.W. Bush, Bill Clinton and George W. Bush administrations. So, unquestionably he has the vantage point and experience to write enough about world and specifically US economics. Him being academically inclined helps this cause well. Compared to the first half of the book that is a bit autobiographical, I found the second half chronicling his time at the Fed and discussing his views on various aspects of world economy much more engrossing.

Until he retired in 2006 he was considered a venerable figure with an extremely steady hand that had guided the US economy through a lot of storms. You can find quite a few books like "Maestro" by Woodward, and "Alan Shrugged" by Jerome Tuccile (remember "Atlas Shrugged" by Ayn Rand?) that will illustrate how the media treated him with awe. In his early years he had hung out with Ayn Rand and has been a true believer of market economics and Adam Smith's invisible hand. So, throughout this book, as a stubborn , he is vaxing eloquent about deregulation and how efficient/self-correcting market forces are left to themselves. But the turbulence the world economy has gone through since the book came out in 2007 makes some of his proclamations sound silly. To quote a passage (page 360), "Rising leverage appears to be the result of massive improvements in technology and infrastructure, not significantly more risk-inclined humans. Late 1950s experience with consumer debt burdens has made me reluctant to underestimate the ability of most households and companies to manage their financial affairs." If some talking head on the TV says this, I can give it a pass. But coming from Greenspan it really makes you wonder how much do we even philosophically understand the way economy works..! If this were true we shouldn't have experienced sub-prime mortgage crisis and the evaporation of Lehman Brothers in the last 4 years.

To quote some more (page 490) he says, "Regulation, by its nature, inhibits freedom of market action, and that freedom to act expeditiously is what rebalances markets. Undermine this freedom and the whole market-balancing process is put at risk." He does have some hand waving lines saying that the invisible hand presupposes that market participants are rational and always act in their self-interest and there could be instances where this is not true. But he keeps arguing that the speed at which monetary transactions take place in the 21st century renders any kind of regulation impossible since the corrective mechanisms won't be able to keep up. But such arguments completely ignore the fact that monitoring individual transactions and allowing/blocking them is not the only way to prevent 2009 type disasters. For example, the Frank-Dodd legislation passed last year in US to maintain better vigil on the financial industry has following provisions among others:

- Creation of the consumer protection agency to oversee mortgages, credit cards and other similar products.

- Asking banks to hold more capital and reduce leverage

- A body to scan the overall economy for threats to financial systems

- Curbing CDO (Collateralized Debt Obligation) type financial instruments that are way too exotic for any meaningful management or making sure financial institutions who deal with them are doing it at their own risk (i.e. they won't be bailed out if things go wrong). 

Such ideas can maintain a leash on the system without blocking innovation itself. I do think that there can be reasonable regulation to make sure playing field is level and companies don't become too big to fail. 

He is really impressed by China's growth but is dismissive of India's since India's growth is still not reaching the masses much due to stifling bureaucracy/corruption. I found his views on energy, US education system, immigration, deficit and social obligations of US Govt (i.e. Social Security, Medicare/Medicaid) very meaningful and in line with what I think is common sense. On the whole it provides an insight into his thinking and by implication the way US economy was being managed for the last couple of decades. Though I won't agree with him on everyone of his views and policies, it is certainly a good read.

Thought Experiment

Came across this piece when a friend forwarded it to me.

http://www.straightdope.com/columns/read/3011/what-would-it-be-like-walking-around-on-a-cube-shaped-planet 

As the URL itself implies, it answers the question "What would it be like to walk around a planet that is shaped like a cube instead of a sphere?". It is a short/neat piece. As you put some thought into even such hypothetical questions, some of the details that come up are pretty intriguing. For example, due to the way gravity will work its magic, the corners of the cube will appear more as mountain peaks to a person standing on the planet rather than corners of a cube. Good mental exercise. :-)

 

Straightdope site seems to have other such pieces as well. Need to explore.

Book Review: Fooled by Randomness

Read "Fooled by Randomness: The Hidden Role of Chance in the Markets and in Life" by Nassim Nicholas Taleb. This is the predecessor to his second book "The Black Swan" I read earlier. I should say this one left me unsatisfied. 

Taleb's main observation is that even the so called experts in various fields, especially in finance, often mistake random occurrences for deterministic cause & effect events. He argues that we as human beings are hardwired to attribute results of our actions to our knowledge and expertise, while in reality they are just random coincidences. It is an important observation. He has divided the book into three parts. In the first part, he looks at the impact of so called rare events (i.e. black swans) and argues that we should analyze all the supposedly successful strategies in the context of entire sample space/human history that will prove practically all strategies to be meaningless. In the second part, he continues elaborating the same idea talking about survivor bias where human beings end up learning only from successes and ignore the failures, thus losing perspective. In the very short part 3, he states that he is also vulnerable to such flaws and is constantly trying to outmaneuver those instincts. 

Some of the examples he has provided are nice easy to understand ones. For example, let us say there were 10,000 investment managers in Wall Street and each year just by luck, half of them make money and the other half lose money and drop out of the business. If this trend continues for 10 years, there will still be about 9 managers who have made money each year in the last 10 years making them look very successful. But since they made money only by chance, their dumb investments exposing them to dangerous black swan events that occur once in a blue moon could wipe them out in one stroke when they happen. But chapter after chapter he repeats this idea in various forms, stories, and anecdotes. He doesn't seem to move beyond this point to explain what else one should do and what are the methods to avoid this pitfall consistently. The basic idea can be explained well in a long article, without the need for a 200+ page long book. While I often wonder if the contents of a book I read could have been succinctly presented in an article lasting just few pages, this time around I indeed came across Malcolm Gladwell's Newyorker article posted  at  http://www.gladwell.com/2002/2002_04_29_a_blowingup.htm that summarizes this well well..! Gladwell's own books suffer from extra fluff as well but his articles are usually well written and convey the central ideas nicely, including this one. The fact that Taleb's investment strategy is to buy options that will bleed him a little money 99% of the days but will make him millions when he hits an unexpected scenario (that usually wipes everyone else out) is better presented in Gladwell's article than in this entire book..! Looks like the hedge fund he tried to run using this idea has been wound up few years back. See http://en.wikipedia.org/wiki/Empirica_Capital 

As I was getting bored, I was wondering if I should give up the book half way through. But kept reading it over a long time thinking there may be more interesting ideas presented further along. But didn't find anything more till the end. In fact after finishing the book looked for a word that describes my inability to let go things incomplete, without pushing myself to finish every task I take up (like reading this book). Quick Google search didn't find me any..! While in general this could be a virtue, I think if I had quit this book early, I might have read two other good books by the time it took me to finish this one. :-)

Bitcoins

I have been following the raise and fall of bitcoin for a while now. If you have never heard of it, here is a quick video clip that introduces the concept: http://www.weusecoins.com/

Ideas of this kind are intriguing since they offer something better than statusquo. But governments are worried that if this gains traction, it may help drug trafficking, arms trading, and such outside the realm of established systems. For those who are uninitiated, bitcoin is a virtual currency proposed by a purported Japanese hacker named Satoshi Nakamoto (which is clearly not his/her real name) that could be created by running a computer program. So, if you like, you can download a program from the weusecoins site listed above, run it on your computer to generate this money. But as people allover the world "mine" this virtual money, the algorithm used to generate it becomes increasingly complex making more money generation difficult. So, by design generation is expected to plateau around $21 million. There are bitcoin exchanges where you can go and pay regular money to buy bitcoins (like you buy stocks) and the price of 1 Bitcoin goes up and down like a stock price. It went as high as about $30 for each bitcoin and then has fallen down to some $17.

 The interesting feature about bitcoin is that similar to cash (and unlike credit card or check based trasactions), once the "money" changes hands, it is not traceable. So, it has created strange bed fellows that include privacy advocates (that are against media and MNC companies tracing all our activities online to target us for more selling/advertising), counter culture population, business people, geeks to drug lords..! Since people can use bitcoins to pay for products and services online without allowing any law agencies or governments to trace them down, it has been a concern for governments.

There have been several concepts proposed on the web that never gained legs. The "internet time" proposed by Swatch could be cited as one such example. Checkout http://en.wikipedia.org/wiki/Swatch_Internet_Time if you have never heard of it. :-) But the idea of software generated currency that is hard to trace seemed to have gained some currency :-) and even has multiple variations now. See http://solidcoin.info/five-reasons-to-own-solidcoins.php for another example. Looks like some version or other of this virtual financial instrument is here to stay. It will be interesting to see if it tuns out to be a boon or a curse.

Protecting home computers from obnoxious web material

I have a cable modem based internet access at home. The wireless router connected to the cable modem serves close to dozen different devices on the home network, which is typical these days for most homes. So, I was looking for an elegant way to make sure content delivered to these home devices are clean.

You can go around installing PC based software for each PC to block hate, pornographic, drug related sites. But then you need to make sure the licenses and individual PC configurations are updated regularly. I don't like loading a lot of bloatware on individual computers. In addition there are no such options for say a gaming console (Nintendo Wii) connecting to internet via your home network.

Next option is to setup filters on the gateway router at home. But home routers do not provide enough sophistication to allow/block all you need.

So, decided to use OpenDNS service for now. To use it, you go to opendns.com and register yourself. Then change the DNS server address your home router uses to look up IP address for various websites to the ones provided by OpenDNS (208.67.222.222 and 208.67.220.220). There are instructions on the site for various routers as to how their DNS entries should be updated. Once you set this up, whenever you access any website, OpenDNS will provide the IP address for all the allowed websites and alternately display a message saying "Access to this site is not allowed from your network" for all the offensive sites. Check out http://www.opendns.com/home for more details.

1. They seem reputable with millions of clients with various media vouching for them.

2. By default they don't track what sites computers in your network access. So, no major privacy concern. But if you want you can turn on logging and see what sites are being accessed from your home network.

3. You can delete the logs whenever you like if logging is enabled.

4. Basic service is free for homes.

5. When you login to their site, you can customize what sites need to be blocked. For example, you can choose to block any one of or all of porn sites, gambling sites, hate speech sites, drug sites, even video streaming sites (like YouTube), social networking sites (like Facebook), shopping sites, phishing/malware sites, etc. You can login and change the settings (for example disable all controls) whenever you like.

I am sure there are other similar services. But this one seems decent. Check it out and let me know if you find it useful.

-sundar.

IBM's Watson on Jeopardy today..!

My son Arjun has been a big Jeopardy fan and a tech geek. So, a highlight in his list of things he is looking forward to is IBM's Watson supercomputer playing in the Jeopardy quiz show on US television starting today on a 3 day run.

If you haven't heard of Jeopardy, it is a quiz show that has been running for decades in US television where usually there are 3 human contestants. Subjects used for the quiz could be anything one can imagine and often involves pun, humor and other language traits. Today, tomorrow and the day after while there will be two human contestants (who are formidable winners in the years past), third place is being taken up by the super computer designed by IBM called Watson. There are a lot of write-ups and news clips on the net about this heavily promoted, highly anticipated show. I am listing just a couple of links here:

http://washingtontechnology.com/articles/2011/02/10/ibm-watson-data-uses.aspx

http://www.hpcwire.com/features/Must-See-TV-IBM-Watson-Heads-for-Jeopardy-Showdown-115684499.html

Please do read them. While on the outside it is designed to be very simple, easy to understand on the show as just a computer taking the place of 1 of the 3 contestants, the audacity of this effort is mind-boggling if you are familiar with natural language processing challenges in Computer Science. In 1997 a previous IBM supercomputer called "Deep Blue" played chess with Grand Master Garry Kasparov and beat him. While it was an amazing feat, chess is much more precise and lends itself well to coding than a challenge like competing against human beings on a quiz show like Jeopardy with full fledged natural language processing capability. Though this proof of concept is via a quiz show, you can immediately see how this could be useful in so many facets of human life. Watson is powered by 10 racks of IBM Power 750 servers, running Linux. It uses 15 terabytes of RAM and 2,880 processor cores, etc. Will see how it fares this week. :-)

Hopefully this will excite a lot of new kids and inspire them to get into Computer Science..! :-)

-sundar.