For fans of Borges, an hour long discussion on the BBC radio program In Our Time.
Scott Kriens, CEO and Chairman of Juniper, speaking at the Stanford Entrepreneurial Thought Leaders lecture series, emphasizes the role of dumb luck in startup success. Stanford students are lucky to have the opportunity to attend these lectures; thanks to podcasting we can all enjoy them. Other lectures in this series I found particularly good (available at the link above): Marissa Mayer (Google), Chong-Moon Lee (Ambex), Carol Bartz (Autodesk), Mark Zuckerberg (Facebook), Jeff Hawkins (Palm). All of these people are smart and have valuable insights to offer.
Another podcast series I recommend is Bloomberg On the Economy, which is a mix of academic and "market" economists (the latter work at banks and investment funds) and the occasional portfolio manager like Bill Gross. I find it very amusing that one can easily find two "market" economists confidently stating completely contradictory predictions on the same day (e.g., right before an employment report or Fed meeting). Sometimes I listen to the podcast a day or two late, so I actually know the outcome already as I listen to the poor guy (sometimes it's a woman) laying out their case for a prediction that turned out completely off the mark. These people must be selected for lack of Socratic self-examination, because the models they play with are so obviously lacking in predictive power, yet they never figure this out. (I suppose it's possible they do realize this, but then they all deserve Oscars for Best Actor or Actress, since they project sincere belief in their predictions.) The academics are typically more thoughtful and, being tenured, aren't under pressure to predict the next tick :-)
I know I'm belaboring the point, but if you forced one of these guys to give error estimates, I am sure you would find them making 3 sigma errors with regularity: e.g., my model says the jobs number is going to be low, only 90k new jobs created last month, with standard deviation of 30k -- gulp, the real number is 180k! And this guy is a Sr. VP or MD at Bear/Merrill/CS/Goldman, whatever. But they never subject themselves to this discipline -- if they did, they'd realize their actual one sigma error is so big that their central value is useless.
You might get the impression I'm only listening to the Bloomberg podcast for comedic value, but I find the reasoning and intuition of the various guests interesting, even if they can't predict anything. At least it helps you understand what a sizeable fraction of market participants are actually thinking.
Finally, if you're interested in recent progress in biotech, I suggest Futures in Biotech with Marc Pelletier (a postdoc at Yale). They cover topics ranging from protein folding to microarrays to neuroscience through interviews with leading scientists. The only problem with this show is that sometimes the real scientist being interviewed forgets the target audience is supposed to be kind of sophisticated and they resort to the usual pop science cliches.
I listen to this stuff when I'm running or at the gym. Thanks to Steve Jobs, instead of fighting boredom or replaying 80's hits from my youth, I can learn something while sweating.
Those market "economists" are amazingly well compensated for being wrong. Come to think of it, so are stock analysts and political pundits. Their compensation makes no sense unless you view them as entertainers.
ReplyDeleteLet's get real here. People who can make correct predictions with any kind of reliability is not going to waste their time being an "economist" or "analyst". They're working for or running a hedge fund or a prop trading desk at one of these banks.
ReplyDeleteEconomists and analysts are hired by investment firms for the "benefit" of their customers only. So calling them "entertainers" is not really so far off the mark.
It's not just that they can't predict - I can't predict the next core CPI number either - it's that they don't seem to realize this. Either that, or they're very happy with their high entertainer compensation ;-)
ReplyDeleteRather than professional actors (i.e., con men, who are plentiful on Wall St.), they strike me more as slightly geeky middle-to-highbrow thinkers who are in love with their little n-parameter models, and truly believe there is value in their curve fitting. I guess I know some engineers who are like this as well. Most of these guys have PhDs, by the way.
Thanks for the Borges link...I wrote my literature thesis at Tech on Borges. One thing I agreed with in the discussion was the comments about the "made-up" works/names in Borges. The amazing thing about these constructions is that they are often quite closely related to real works and are almost always intimately tied to a major theme of the story, often in a quite complicated manner. Often you'll find that a character in one of his stories has the first and last name of certain obscure philosophers who have said important things about the topic Borges is discussing. Very cool if you want to spend the time digging around the stories. I did however disagree with the panelist who dissed the intellectual depth of Borges work. Of course my literature thesis was on Borges and the Godel incompleteness theorem, so you can guess I have a diffrent perspective!
ReplyDelete