April 29, 2004
If I had seen more significant digits in the earlier news articles, I might have noticed:
The exact offering, $2,718,281,828, is the product of "e" and $1 billion, where "e" is the base of the natural logarithm--a logarithm especially useful in calculus--and equals 2.718281828.... [CNN]
I am geek enough to enjoy that. Hell, it's at least as logical as any other IPO valuation I've heard of in the Internet age. Actually it's deeply logical, in that it recognizes the fundamental irrationality of IPO prices. There's no pretense. And is it an accident that e is, technically, an irrational number? ;)
Frankly, the document they released today dubbed "An Owner's Manual' for Google's Shareholders" makes me feel more positively about Google's prospects than anything I have ever heard from or about them. It says things like:
"Many companies are under pressure to keep their earnings in line with analysts' forecasts. Therefore, they often accept smaller, but predictable, earnings rather than larger and more unpredictable returns. Sergey and I feel this is harmful, and we intend to steer in the opposite direction." [Another CNN article; definitely worth reading]
And, about their decision to have two voting classes:
"While this structure is unusual for technology companies, it is common in the media business and has had a profound importance there. The New York Times Company, the Washington Post Company and Dow Jones, the publisher of The Wall Street Journal, all have similar dual class ownership structures. Media observers frequently point out that dual class ownership has allowed these companies to concentrate on their core, long-term interest in serious news coverage, despite fluctuations in quarterly results." [Same CNN page]
I find this very interesting. They seem to be trying to engineer a solution to the corporate problems other companies have experienced, just as a good engineer addresses technical problems that earlier projects have experienced.
And the best engineers recognize that sometimes, they have to fly by the seat of their pants. And the very best engineers know when they have no choice but to do so; they don't pretend they know exactly what they are doing in such a case. They have integrity. They don't hide the facts. Is the choice of an irrational number for the IPO valuation a conscious wink to those who might care, saying this is one such case? Is it a statement: "We have the integrity to act according to our principles, even to the extent of making choices that are not what Wall St. would necessarily understand or approve of in the short-term"? I hope so.
Maybe this will turn out to be as interesting an experiment as it seems it has the potential to be. Cool!
Google files for IPO
At $2.7 billion, not the $25B people have been speculating it's worth. That makes more sense. Google does have a lot of eyeballs at this point, and there is a chance that it can convert many of those users into a situation where they are totally dependent on Google for their email or other services. Services like email are sticky in the sense that if a user has his entire email history ensconced in Google's servers, it will be hard for even Microsoft to get that user back as an email customer. If Google is successful in converting huge numbers of users into that kind of sticky situation, rather than the current situation where they'll defect as soon as a better or more convenient search engine comes along, then it could conceivably be worth $25B.
There's probably about a 10% chance of that happening, where the main risk is the fact that Microsoft is fully capable of creating a search engine of equal quality that is more conveniently accessible within the Windows GUI. And doing so before Google has converted many users.
Gmail has opened my eyes to the possibility that Google could be successful in "stickfying" large numbers of users; that's why I'd give it a 10% chance rather than a 1% chance. I didn't think their blogging software was enough, or their social software (Orkut). But Gmail has a chance.
So a 10% * $25B IPO doesn't sound absurd to me at this point. That's not far from $2.7B.
If a lot of suckers pay enough on IPO day to bring the first-day valuation to much more than that, some smarter people are going to make a lot of money by selling on the second day, I believe. ;)
A great math joke
April 28, 2004
Handling Redundancy in Email Token Probabilities is a new article by yours truly about the chi-square-based technique for spam detection. (That's the PDF, here's the Postscript version.) The abstract:
One of the many techniques which has recently been employed for filtering spam is one describedin the Linux Journal article A Statistical Approach to the Spam Problem. This technique incorporates ideas from the seminal article A Plan For Spam as well as R.A.Fisher's technique for combining p-values by means of the chi-square distribution. The technique presented in here takes the chi-square-based approach a step further by taking into account two facts: a) there is redundancy in the token probabilities, and b) spam andham emails have different amounts of such redundancy. Fivefold cross-validation was carried out on the new technique and is described here testing whether these factors actually lead to better performance. The results were positive and statistically significant.
Note: the link to the article above is to version .94; earlier versions are still available still here , here, here, and here, but they have some typos, and the original has an error in Eq. 3. Many thanks to Greg Louis and David Relson and others for pointing them out. Also, Laird Breyer has pointed to areas where I could have done a better job of explaining things, so that has also led to improvements in the new drafts and is appreciated. You should definitely use version .94.
Cleaning up Python __init__() methods
The ASPN site yesterday presented a technique to eliminate a lot of verbosity in Python __init__() methods where a bunch of arguments to __init__() are assigned to instance attributes. Such assignments are a frequent source of error due to forgetting one or more assignments and/or making spelling errors for the names.
They presented a function for doing the assignments automatically, but required that the __init__() that uses it has no logic with temporary variables -- an unworkable restriction IMHO.
So I added a comment with a version that works even when there is other logic with temporary variables. If you're a Python coder you might want to check it out.
April 27, 2004
Lots of movie downloading
Ipsos-Insight's quarterly digital-music study Tempo reveals that 21 percent of American downloaders aged 12 and over have downloaded a full-length motion picture.
The research reveals 37 percent of 18- to 24-year-olds have downloaded a full-length motion picture from the Internet. Men are more than twice as likely than women to acquire a digital motion picture online. [PCWorld]
A full-length movie is a big download. 21% is impressive.
April 26, 2004
The OpenOffice.org suite now claims a 7 percent share of the midsize business market, according to Jupiter Research, the research division of Jupitermedia Corp., in Darien, Conn. [eWeek]Wow, I'm impressed. That's much more than I would have guessed at this early stage. And I'm one of those who take it as a given that the percentage is likely to continually grow.
Online technology company CNET Networks Inc. (NasdaqNM:CNET - News) on Monday launched a free digital music service, allowing users to search and download what it said were thousands of songs contributed by independent and unsigned artists. ... The music download service will be separate from the MP3.com Web site, which CNET has said it will relaunch soon as a music information site. ... The former MP3.com, which CNET acquired late last year, offered similar functions to those in the new music download platform. Many artists and fans of the site complained when CNET closed it down, fearing that hundreds of thousands of songs would be lost permanently. ... CNET said it plans to add new technology and community features to its download service, which it intends to be the largest free music download platform. over the course of the year. [Reuters]
April 20, 2004
Why condemn Rantisi killing? I don't get it. Honestly.
U.N. Secretary-General Kofi Annan said the killing of Rantisi was a breach of international law and called on Israel to end its assassinations.
"The only way to halt an escalation of violence is for Israelis and Palestinians to work towards a viable negotiating process aimed at a just, lasting and comprehensive settlement," Annan said. [CNN]
I find this impossible to understand. The guy was head of an organization that sent suicide bombers into Israel to murder Israelis. The guy must have personally ordered such attacks or at least given the go-ahead.
Why should Israel not go after someone who is doing such things? He was escalating the violence by ordering or approving the murder of innocent civilians because he thought their deaths was helpful. Why should be be allowed to continue doing as many as he wants of such acts?
It just makes no sense to me. If you have a way of making sense of it, please let me know.
April 16, 2004
Manipulating a certain resource in a good cause
Copying taint.org's manipulation for a good cause: jew. (To understand what the good cause is, see what the dominant search engine's first hit in response to that word, and where my own link goes. And consider doing the same thing on your blog.)
I'm not a jew, but my wife and kids are, and I've read enough about World War II to understand what a bad place the world can be for good people.