262ventures.com

A recent blog entry from the Long Tail discussed fragmentation of the search engine market, specifically about the trade off between breadth and depth of quality search. On one hand Google crawls the entire web and determines relevancy algorithmically covering a broad space. On the other, there are a bunch of “human-powered searches”, such as Squidoo, who use human interpreted contextual analysis (read: reading with your eyes and writing at your keyboard) to create more depth and relevancy for more common search terms. The Long Tail appropriately described it as “leaving the long tail to algorithmic search and seeking success by relentlessly editing down to a human-edited short head”.

This may be a good strategy if you can achieve strong market share. This seemingly works for Alltop, Guy Kawasaki’s feed sorter. But allow me to invoke the story of Looksmart in the 1990s. Their human ‘categorization’ method, which consisted of my buddy Dave sorting hundreds of websites per day into their appropriate category, was a rough approximation, a web 1.0 version if you will, of Squidoo.

In this age, I’m not sure anyone doing search would be willing to not use Google. It’s the gold standard for long tail searches. But it’s pretty damn good at the common stuff too. They took charge of an industry long before Looksmart even went public.

And you can ask Dave how that good old 24:1 reverse split worked out.

I am not qualified to make any economic forecasts. But I have a gut feeling that over the next decade or so, we will see a substantial movement of left-ish moderates begin to embrace the financial modus operandi of the right. I think there will be a nationwide push to reduce tax burdens. Times are tough right now. If you consider some major factors such as inflation, devaluation of the dollar, the ineffectual fed rate cuts, and toss in a some foreclosures, it looks like you’re gonna have a witch’s brew of stagflation for the upcoming years. And if the belt gets squeezed beyond the last belt-hole, you might get this.

So we’re stuck right? Can’t tax more, yet we’re mortgaging our nation away. I think the way we get outta here is to reinvest in innovation, talent and education before we get to the point that we’re Europe. Let’s get the core of the US excellent again.

Awesome!

I am looking forward to Boston ‘09.  I missed Patriots Day dearly yesterday.

Today is earth day.  Go out and plant a tree or walk to work instead of driving.  Better yet, come up with new innovative solutions and technologies that will help out towards addressing our energy, water, trash, recycling needs.  We’ve still got a long way to go and implementing new technologies is the way to achieve scale.

Whatever you do, don’t give your credit card information to the ‘volunteers’ on the street who ask you ‘if you have a few minutes for the environment.’  I can’t think of anything more value erosive than to give a donation to an organization who pays ‘volunteers’ to collect donations.  At best you are getting a 1x return on your contribution and in reality you are probably giving 10 cents per dollar towards end usage.

Instead I would spend that money educating kids on environmental stewardship.

This is a huge event in Boston, with sideline cheerers often 8 deep along Commonwealth avenue. It’s augmented with an early Red Sox game and a statewide holiday [Patriots Day]. Lots of early partying, music, and general merriment. My buddies Blaine and Bernat are running this puppy and at the 10K, it looks like Bernat is outpacing Lance Armstrong. Good luck guys.  I’m looking to join them in 2009.  We’ll see if we can actually wind up putting the 3:10:XX qualifying time together.

Back at MIT, we would get Patriots Day off and have a ChemE department intramural team for the octathon. This event would consist of 8 events [basketball, softball, flag football, volleyball, soccer, water polo, ultimate frisbee, and ice hockey] played over the span of the weekend. Those were great times. Happy Patriots Day everyone.

Erik passed along a link to an HBR podcast interviewing HBS professor Rakesh Khurana’s take on whether professional business schools should implement a “Green Hippocratic Oath.”  The basic idea is to establish a codified set of principles whereby professional business managers seek to maximize corporate and social well being, as opposed to money or profits, through its practices and processes.  Professor Khurana’s view is that business school teachings have settled on the notion that students go to business school to develop an elite contact list, obtain a job that pays handsomely and maintains a reasonable moral balance, and then retire and give back after accumulating a large treasure.  He continues to argue that concepts pervasive in business school education such as “maximizing shareholder wealth” or “competitive advantage” need to expanded to include “social welfare” and “consideration for environmental impact.”  One manner this can be accomplished is through a green oath.  There are eloquent arguments for and against such an idea listed in the comments section of the post.

Two points.

1.  I personally really like this idea.  Dan Ariely, in Predictably Irrational, empirically demonstrated that the effect of an ‘honor code’, a set of rules that fall in line with societal expectations, is a strong reinforcer of behavior.  The obtainment of a business management  ‘profess’ional degree, as Ariely writes in the book, should require the practitioner to publicly profess their responsibilities as stewards of business and society.

2. I like the concept of integrating ’social good’ and other values into the utility function.  Let’s call ‘well-being’ a proxy for money, social welfare, positive environmental impact.  Economists have long known that every additional dollar made has monotonically, decreasing added value.  If you take the mathematical extrapolation of “maximizing profits” to its limit, there is zero added value created.  Clearly the outmoded concept of making a ton of money at the expense of all else is the equivalent of being stuck in a local maxima.  Global value [both in a geographical and parameter sense] can be created without being realized only in US dollars [which aren’t that valuable these days anyway].  The practical implementation is obviously completely unclear but the debate and attempts at determining this is worthwhile.

A few months ago, I was up at Stanford University, sitting in on a few of the E-Week events.  It included an interesting forum regarding ’social entrepreneurship.’ The forum consisted of the three founders of B Labs, whose goal appears to be the branding and advancing of a new type of corporate structure (B Corp vs. S Corp or C Corp).  It is a thoughtful and earnest attempt at creating a ‘badge’ or ‘honor code’ whereby corporations can act in a manner to increase well-being.   Check it out.  Oh and these guys are responsible for the And1 shoe company, so at least they were responsible for creating highlight reel dunks over the years.

I recently revisited an idea I’ve been playing with for some time - the wedding seating optimizer. The idea is that everyone spends an inordinate amount of time planning the seating chart for their wedding - Aunt Bea and Uncle Joe are divorced and can’t sit together, we want our best college friends to meet, etc. - and couldn’t a computer do this for you instead? At the core of the wedding seating optimizer is an equation that calculates how good or bad a particular seating arrangement is, which I’ll call the cost function. Like golf scores, lower cost function scores are better. The program implements an algorithm that seeks to minimize the cost function with an optimum arrangement of guests.

To test the wedding seating optimizer, I created a fake wedding with 100 guests, added the aforementioned Aunt Bea and Uncle Joe to the mix, and started the thing running. I set out 10 tables, and I got the following results:

Table Number of Guests
1 99
2 1
3 0
4 0
5 0
6 0
7 0
8 0
9 0
10 0

Oops.

What happened here? Quite simply, I neglected to incorporate any information about the optimum number of guests per table into my cost function. And since everyone at the wedding (save for Aunt Bea and Uncle Joe) gets along, we ended up with one big table (and Uncle Joe crying alone over his wedding cake). More generally, the problem was a case of “Too Much of a Good Thing.” When I first wrote an equation for the cost function, it never occurred to my limited human brain that one could try and seat so many people around a single table. I was limited by my mental image of a wedding table with 8-10 people around it, and I assumed from that vantage point, that adding one more friend to a table of college buddies would be a good thing. Clearly, the computer is not so mentally limited (or, if you prefer, is too stupid to know that 99 people can’t sit at a single table), and so it found the best possible solution to the seating problem I laid out.

“Too Much of a Good Thing” is a common and insidious problem in any numerical optimization scheme, and arises simply because the cost functions we propose are frequently too simplistic. It’s so prevalant in fact, that I would propose the following law:

If you fail to account for “Too Much of a Good Thing”, you will inevitably receive “Too Much of a Good Thing”.

More importantly, “Too Much of a Good Thing” is also the cause of many real life problems, frequently in the form of bubbles. Economists are fond of calling this “The Law of Unintended Consequences”. Take for example, the recent housing boom and bust. It’s tough to spell out society’s housing cost function, since the cost function would incorporate our collective opinions on housing. But, a simple version of our cost function might be described by the conventional wisdom “Home Ownership is a sound investment promotes positive social effects”. The U.S. Government promotes home ownership though tax breaks. Social norms promote home ownership as a positive good. And in the down stock market of 2000-2002, a real asset such as a home seemed like a good idea. The state of the U.S. in 2001-2005 conspired to create a situation in which “Too Much of a Good Thing” (in this case, home ownership) was ignored, and a bubble was the result.

I have had the chance to listen to “We Appreciate You”, the soon to be released final(???) album from The Plastic Constellations.  WOW!!!!!

For those of you unfamiliar with TPC [check out the previous post about Pandora], they are a Minneapolis/St Paul based hard-charging, post punk, inclusion rock band who have been playing together since they were in junior high school.  Jeff Allen, one of the two front-men of TPC, also happens to be Erik’s brother.

In any case the new album is completely kick ass!  Their music has always burgeoned with energy, interesting melodies, slightly atonal and syncopated counterpart, and yell-with-ya choruses, but this album has got all of that in spades.  I’ve followed the progression of their music from Mazatlan to Crusades  [with bits of stuff from Let’s War!].  With We Appreciate You, I am instantly made aware of their honed tight composition, high fidelity sound, low fidelity spirit, and visceral musicianship.  As they’ve self-announced a hiatus for the band, the latest and potentially last album from TPC sounds of a cathartic culmination of farewells.  To me, it seems like TPC is only getting better with time and it’s a shame for us fans, left dizzy and wondering, what these guys were to be capable of over the next few years.
If you like amazing live shows, lots of pogo-ing, I highly suggest you get on a plane and check out this show.

Top 3 songs thus far: Phantom Canyon,  Disastrophe, and Perched on a Porch.  If you like their released singles “Stay that Way” and “Black Market Pandas” (check these out here) you’re in for a serious treat with the rest of the album.

Kim, a couple friends of ours (Ban and Hans) and I ran the Agoura Hills Great Race this past weekend. It was a family-oriented outdoor festival with dozens of vendors, free food, a climbing wall, and jumping castles. It also included a half marathon, 10K, 5K, and family fun run.

Kim and our buddies ran their first 10K in the picturesque but hilly trails of Agoura Hills. I’m super proud of them as we’ve been running monthly 5Ks. They all finished for PRs and personal goals. Hans wanted to finish in under an hour and squeaked in there by 3 seconds! Way to kick at the end Hans!

I ran the half marathon in 1:41:XX, about seven minutes off my PR. But considering the thousand foot climb that peaked at mile , I’ll take this one as a victory. It was the first half I had done in about a year and my goal this year is to get that PR down to the 1:30:xx range. I think it’s doable as I felt incredibly good after this race, in contrast to the Brooklyn half last April. I guess having trained for a marathon in between really made a difference… I would’ve never thought that 13.1 miles could be a (relatively) short distance.

Oh… and I made the video (wearing the old school blue CTR jersey)!

Mark Sweeney of the NY Times has an article up about how Tiger Woods achieves his success (of course other than that sweet Stanford education). He builds a statistical model (probably through linear regression) to determine the parameters that are most predictive of achieving a low score. And sure enough the heavily weighted parameters (greens in reg and putting) are among those which Tiger is best.

It is indeed quite cool that folks are using stats to track and improve performance on the PGA. Boy we would love to get our hands on some of that data.