Tuesday, May 31, 2005

Entropy, creationism, and intellectual conservatives

Prof. Myers' title for a recent article at a web site called, "Intellectual Conservative" was just so funny that I had to write this "me too". His title is "Babu G. Ranganathan: goddamned moron". The article starts off:

Very few scientists have considered or pondered the implications of the law of entropy upon the theory of evolution.

This is the equivalent of saying, "Very few scientists have considered or pondered the implications of centrifugal force upon the theory of a spherical earth." Stupid in so many ways, it's hard to know where to begin. At least we have a concrete definition of what "intellectual conservative" means.

Topics: , , .

Friday, May 27, 2005

Mad enough to switch

There's a very interesting thread over at the Security Awareness Blog (linked from title). Will Schwartau decided to switch his company from Wintel to Macs, and in a series of postings, entitled "Mad As Hell", he details why and how the switch is going. A good part of his reason for switching is hardware (and hardware support): hardware vendors are notorious for blaming the software (funny that this seems especially common when the software is written by someone else). A choice quote:

The cost - the unacceptably immense financial premium that the WinTel world demands of its users - is going to be the straw that breaks the WinTel back! I could no longer afford WinTel.
I know that there are a lot of rumors about Apple and Intel going around, including the licensing of the Mac OS to third party computer makers. Mr. Schwartau's comments argue very clearly that the Mac experience is not just about software -- the hardware and support are as important. I don't see why any end user would care if their Mac's CPU was Intel, but cheap architecture, buggy BIOSes, even generic PC cases will destroy the Mac's image (and it's reality) as a vehicle for a high-quality, even emotionally compelling, experience.

Want more of an argument against Mac cloning? Consider recent comments by Paul Otellini, Intel's new CEO, when asked by Walter Mossberg, "Why are Intel's demo products always cooler than the actual products its customers make?"

Part of the problem is today the desktop [computer] business is a zero margin business for many of our customers.
Apple's computer business is not zero margin, and all the market share in the world is not worth trading for that.

In that same Wall Street Journal article, Mr. Otellini recommends Macintoshes for people who want a computer that just works, at least as far as security is concerned.

Tuesday, May 24, 2005

What kind of Homeopathic remedy are you?

In the spirit of of web tests such as the Star Wars Personality Test, I followed a link from Anne's Anti-Quackery & Science Blog to ABC Homeopathy (linked from title). There, I took a little test to explain what ails me. I chose whatever interesting symptoms I found, plus whatever struck me as funny (be forewarned, my sense of humor never progressed beyond the potty stage). You can do the same thing, and learn what kind of homeopathic remedy you are. It's at least as good as any of the other on-line personality tests.

I said my problems were: mental (dazed, impaired intellectual faculties, insecurity, anxiety, fearful at night, fearful while having bowel movements, tired, and dizzy), throat (mucus, expectorating balls at night), stool (bloody, greasy, and slimy BMs), and perspiration (with diarrhea and during BMs), plus some general ones (anxious, flabby feelings, especially at night or during siestas). I learned that:

I am Sulfur
Sulfur is good for many things, including the mind, vertigo, head, eyes, ears, nose, face, mouth, throat, abdomen, rectum, stool, urine, genitals, respiration, cough, expectoration, chest, back, extremities, sleep, chills, fever, perspiration, skin, injuries, and general complaints.

Monday, May 23, 2005

Do we really want more students interested in CS?

The Chronicle of Higher Education has an article (linked from title) summarizing the latest statistics on the number of CS majors and incoming freshmen expressing an interest in majoring in computer science. They're both down: 32% and 59%, respectively, in the last four years. The main theme is that this is a bad thing. In the opinion of industrial and academic employers, that is. Seems companies will need to "work harder" to get good employees, and that's a bad thing. As Richard F. Rashid, senior vice president for Microsoft Research says:
It's a major concern for us because we're a company that runs on people. Our hiring has continued to go up, but unfortunately what we're seeing right now is a decline in the potential supply.

From an employee's perspective, fewer people seeking jobs is a good thing. And, frankly, while having lots of graduate students to shovel code may help some university research, I'm not convinced that most of the system building that results is truly significant. Interesting, yes, even neat. But not likely to have any significant impact. University faculty should have their students' interests in mind when talking or writing about the job market, and I'm not sure we do when we talk of declining enrollment as a bad thing, or, even worse, a crisis. Declining enrollment is a rational response on the part of students to a significant drop in the job market.

The most common opinion is that enrollment swelled during the late '90s due to the dot-com bubble, when a CS degree was viewed as a ticket to wealth. Do we really want to fill our classes with students whose primary concern with CS is the money they think they'll make? Isn't it better to have fewer students who truly love what they're doing? It's certainly better to teach them, and I believe those are the people who will be happy in their chosen careers.

There are of course other potential reasons for the decline. For one thing, there has been a marked increase in enrollment in IT programs, which focus more on the nitty-gritty of running systems than on theory and design. To the extent that this allows students interested in computing not as designers but as more infrastructure-oriented team members, this is great. The challenge is to make an IT degree something other than "CS lite". The other major reason given is declining standards in mathematics. It's traditional for each generation to call the next a bunch of spoiled underachievers, and I don't know how much this is true here. From personal experience with my own children (in early elementary school, right now), I see pluses and minuses with the current educational system: a fair amount of structure and rigor with better integration of concepts such as algebraic thinking and logic on the one hand, but on the other a standardized-test-driven approach that basically means cramming the "gist" into the kids' heads just before the exams and then coasting for the remainder of the year.

There's a table in the article that illustrates very clearly what the demonstrable problem in the field is. It shows the percentage of freshmen expecting to major in CS, starting in 1990 with 1.7%, increasing to 3.7% in 2000, and then dropping to 1.4% in 2004. This seems like clear support for the thesis that there's trouble in River City. But, look at the breakdown for men and women: 2.3% to 6.5% to 2.8% for men and 1.1% to 1.4% to 0.3% for women). This is even more troubling when we recall that almost 50% of CS majors were women during the early to mid '80s. So, the problem isn't that too few freshmen want to major in CS -- the problem is that too few women want to.

What is repelling women (and we have to put it this way, because it wasn't always so)? Two hypotheses seem to be discouragement of young women from mathematics at an early age and the generally miserable reputation of the computing workplace. I can't comment on the former, other than to say that outreach and greater cooperation between computing professionals and K-12 teachers should help. The response to the latter seems to be "we need better PR". Again, Microsoft's Rashid:

You need to talk about the romance of the field. It's not all about people sitting in cubicles eating pizza and typing away endless hours on a keyboard.

There has always been a "macho" subculture of computer professionals who take pride in the number of hours they put in. The dot-com boom ingrained this into popular culture. Merely talking about how it's not "all" about endless hours behind a keyboard doesn't address the issue that this is still the expectation. I have a radical idea: how about Microsoft leading the way in instituting a real 40-hour work week? How about Microsoft getting rid of the practice of hiring "temporary" technical staff? They could corner the market on new graduates. I'm being naive, of course -- hiring lobbyists to get Congress in increase immigration quotas is much more cost-effective.

Wednesday, May 18, 2005

Tuesday, May 17, 2005

45 and Too Old to Work?

The title is from (and linked to) a Ziff Davis article about the difficulty that more experienced tech workers can have finding a job. He cites a "friend" who at first had no luck job hunting, then had much better luck when he reduced his age to "indicate that he was in his 30s". The author doesn't understand why companies would prefer to hire younger, less experienced workers who are less likely to keep the job for a long time. I'll venture some hypotheses:

  • They can be paid less.
  • They will happily work longer hours (tech ennui hasn't set in yet).
  • They have fewer family commitments (less time off/lower benefit costs).
  • They are less independent/easier for their bosses to control/less threatening to their bosses (who, after all, are likely the ones doing the hiring).
  • Companies don't want to "over hire": they prefer mediocre employees who will be happy with their jobs than someone "over-qualified".

Sunday, May 15, 2005

Best podcasts?

In this posting, I'm going to answer what I'm sure is one of the questions uppermost in your mind: what podcasts am I listening to? There's all sorts of podcasts out there, and I tend to listen to the geeky ones. Unfortunately, most of the people who make them tend to spend the first 5 or 10 minutes on title music, sound effects, and various personal stuff that I don't care about. They also say things like, "the wife," which for some reason annoys me. Anyway, there are three podcasts which stand out for me, enough so that I thought I'd write about them.

  • The first one is Science @ NASA, which is a great website, and which has a number of delivery vehicles, including podcasts at the feed: <http://science.nasa.gov/podcast.xml>. The audio quality leaves something to be desired, but the segments are short and sweet.
  • The next is Morning Stories from public radio station WGBH in Boston, with a podcast feed at <http://streams.wgbh.org/podcast/morningstories.xml>. An eclectic collection of short shows with very high production values.
  • And finally, there's In Our Time, from BBC Radio 4. Excellent, in-depth shows that take the time to seriously cover a wide range of topics. The host, Melvyn Bragg, sits down with three guests each week to discuss topics in history, science, religion, culture, and philosophy. You can find the podcast feed at <http://www.bbc.co.uk/radio4/history/inourtime/mp3/podcast.xml>.
All three of these shows are part of highly informative web sites with much additional information and multiple ways of obtaining shows besides RSS feeds. I suggest you skip one of your usual blog reads today (after you're done with mine, of course) and go to one of these sites instead.

Friday, May 13, 2005

These are the voyages...

Time for something lighter to comment on rant about. I just finished watching the final Star Trek: Enterprise episode, "These are the voyages...". This was written by Rick Berman and Brannon Braga, the two gentlemen who brought the franchise to a premature end with their poor writing and incomprehensibly bad judgment. So, what would you write if you had the chance to end a Star Trek series?

Well, I may not be a fiction writer, but it seems to me that you'd want some emotional impact. What do Berman and Braga do? They do just about everything they can to suck all of the emotional impact out of the episode. First of all, they distance the audience emotionally from the plot and characters by having them all be just a programmed re-creation on the Star Trek: The Next Generation Enterprise-D's holodeck, viewed mostly by a smirking Riker as a means for him to decide some throw-away decision that's supposed to be hard to make but isn't presented in any way that makes the audience give a damn. Second, they break the action into little pieces as Riker repeatedly halts the simulation to go talk to Troi about this or that. Hell, technically, this isn't even an Enterprise episode, it's a strangely depopulated Next Generation one (sans Picard and the other recognizable cast members save Riker, Troi, and Data's voice).

Then, they commit what must be the number one cardinal sin of fiction writing: if you're going to kill off a major character, it's probably not a good idea to have someone mention it in passing ahead of time. Basically, the audience is just left wondering -- in an intellectual, rather than emotional, way -- how Trip will buy it. And when he does, not much time is spent on it. There's a little bit of sentimentality displayed by the characters afterwards, but they get over it quickly.

I won't even go into the internal inconsistencies, which were mind boggling. Suffice it to say that six years had supposedly passed from the preceding episodes (another way to reduce emotional impact), but nobody had changed, everyone had the same job, same rank, same clothes, etc. I guess Friday the 13th really was unlucky, for us fans.

Well, I feel better getting that off my chest. Let's hope there's no more Star Trek until Berman and Braga are retired for good. Sadly, we've learned that bad Trek is worse than no Trek.

Thursday, May 12, 2005

Welcome Skeptics

The eighth Skeptics' Circle is on line at Pharyngula, linked from the title above. For those of you who came from there, this post conveys no useful information (very different than most blog posts), so welcome.

Wednesday, May 11, 2005

Yahoo Introduces Online Music Service

OK, so Yahoo introduces an online music subscription service, and Apple stock tanks by nearly 10% before recovering some (as I write this, it's only down about 3.5%). A couple observations of mine:

  • The reason Apple's stock went down is concern that the subscription approach will cut into iTunes music store sales. Is there any evidence for this? Do movie rentals cut into DVD sales? The linked AP story also says, "Many record label executives prefer the subscription approach, Leigh said, because consumers are more likely to sample songs from relatively unknown artists, a phenomenon that helps the industry create more moneymaking stars." I believe the part about the record companies preferring subscriptions, but not for the reason they give. They prefer it because it generates an ongoing revenue stream, while music sales require continuous generation of new, desirable content.
  • Part of this announcement is the new platform war. Many software vendors produce Windows-only, because they feel the Mac's (or Linux's) market share doesn't justify the expense of an additional version. Here, however, you have Yahoo (like the other subscription services) targetting Microsoft-DRM-only players. In other words, they are building a business model that ignores the dominant portable digital music platform, the iPod. The only rationale behind this would be the expectation that the iPod's market share will drop significantly, and fairly soon. Note that there's no technical reason, other than choice of DRM system, for excluding the iPod -- Yahoo's software could deposit downloads into the iTunes software's library (like podcast software does), and the interaction with the iPod would be pretty seamless. Since people who prefer to buy and own, rather than rent, music will have no incentive to go with Yahoo, I see Yahoo splitting the non-iPod subscription market, rather than taking market share from Apple.

Tuesday, May 10, 2005

Grading by the pound

Continuing on the theme of formulaic evaluation of papers is the article linked from this post's title, which describes evaluation of the new SAT essay. Apparently, the motto for this part of the exam must be "bigger is better". Better than getting the facts correct, even. Consider this quote of Les Perelman's (a director of undergrad writing at MIT):

He was stunned by how complete the correlation was between length and score. "I have never found a quantifiable predictor in 25 years of grading that was anywhere near as strong as this one," he said. The shortest essays, typically 100 words, got the lowest grade of one. The longest, about 400 words, got the top grade of six. In between, there was virtually a direct match between length and grade.
One might think that the College Board is using exactly the same software to do their grading as I blogged about here.

Meanwhile, the downward slide of SAT verbal scores continues, noticeable despite the attempt by the College to obscure it by renormalizing the scores in 1995.

Computer grading of student writing

Click on the title to see a USA Today article on software used to automate grading of student papers. Then consider what a colleague of mine has to say about the crappiness of the grammar checker in Microsoft Word. A choice quote from the USA Today article:

When the University of California at Davis tried out such technology a couple years back, lecturer Andy Jones decided to try to trick e-Rater.

Prompted to write on workplace injuries, Jones instead input a letter of recommendation, substituting "risk of personal injury" for the student's name.

"My thinking was, 'This is ridiculous, I'm sure it will get a zero,'" he said.

He got a five out of six.

A second time around, Jones scattered "chimpanzee" throughout the essay, guessing unusual words would yield him a higher score.

He got a six.

Maybe this is what the WMCSCI people use to referee papers.

Intelligent Design's argument from ignorance

I'm very much struck by the observation in the blog linked from the title above that the bulk of the "Intelligent Design" (ID) creationists' arguments are actually arguments from ignorance. Basically, they say things like, "Nature is so complex; I can't see how this complexity could have occurred in the absence of purposeful design by some supernatural being." An argument from ignorance says that, since I don't know or understand something, it must be false. Clearly, this is nonsense: there are many things I don't understand, but they don't require my understanding to be true. The ID creationist argument is entirely equivalent to saying, "I can't imagine why anyone would want to fly an airplane into a building, and therefore 9/11 must not have happened."

Saturday, May 07, 2005

Home Science Tools: Not!

When are home science educational materials not science educational materials? When they're marketed by a bunch of religious crazies, masquerading as science education. Let me explain. In my family, we try to supplement our children's education at home. One of the things we want to do is to give them an appreciation of science as fun, and this is pretty easy, since we're all born scientists: we have lots of questions and we enjoy performing experiments (such as, "How many times must I drop this food on the floor before one of my parents explodes?"). Some of us are fortunate enough to have this reinforced when we're small; others have it pounded out of them and grow to regard science with suspicion, confusion, or disdain.

So, anyway, we are always on the lookout for educational materials for our kids. We recently received a very slick looking catalog from a company called "Home Science Tools". At first glance, it looked impressive. Lots of microscopes, prepared slides, glassware, nature books, dissection kits and specimens, telescopes, rock collections, chemistry sets, you name it. Then I come to the "science" curriculum kits. The first thing that alerted my suspicious nature was that one of the curricula was named "Bob Jones". OK, that might just be a rather unfortunate fact of life -- the publisher happens to have that name. So I turn to the curriculum section of the catalog to look at the books in detail, and my fears are justified. Here are some of the titles they carry:
  • Exploring Creation with Astronomy/Botany/Zoology/General Science/Physical Science/Biology/Chemistry/Physics/Marine Biology. Except for their titles, there is nothing in the description to make one believe that these aren't legitimate books.
  • A curriculum from Bob Jones "University" Press, which "...provides and excellent science curriculum that is very thorough and recognizes God as the Creator of all things." I believe it also promotes regular bowel movements.
  • Advanced Physics in Creation. Apparently, there is no conflict between learning nuclear physics and disbelief of isotope dating techniques.
  • The "God's Design" curriculum, including, Our Planet Earth, which covers plate tectonics, I suppose either to say that it doesn't exist or to assert that those plates are zipping along at breakneck speeds, so as to be able to have moved an appreciable distance during the 4000 or 6000 or whatever number of thousands of years that the kooks believe is the age of the earth.
  • "Media Angels" unit studies, including Creation Anatomy, Creation Astronomy, Creation Science, and Creation Geology.
  • An entire section on "Creation Science", with books on the great ice age and flood, creation vs. evolution (let me guess: creation wins), The Grand Canyon Catastrophe DVD (that must be some awesome footage), and Refuting Evolution, by Jonathan Sarfati, Ph.D., which "...was written to expose the flaws and misinformation in Teaching About Evolution and the Nature of Science, by the National Academy of Sciences" (one supposes that Dr. Sarfati could kick most any NAS member's ass).
  • An entire "Intelligent Design" section, with all the books you've heard about lately and then some.

I could go on, as there's much more, but frankly, I'm a bit nauseated by it all. They're well on the way to completely integrating "science" and their brand of religion. The medieval church would be proud of them. They've got more pseudo-science books for home-schoolers and home education than all of the real science books written for that purpose that I've seen. And these folks aren't the only ones out there peddling this crap; they just seem to be the slickest and, what is to my mind the most troublesome, the most subtle. If I go to a web site and it says right out front that it provides supplies for Christian home-schoolers, then I know to move on. But, you really need to examine these folks' advertising carefully, reading all the way to the back of their catalog, to know the real story. So, caveat emptor. Now, I need to write them a smartass letter and ask to be removed from their mailing list.

Friday, May 06, 2005

WMCSCI Update

Jeff Erickson (who coined the term "spamference") has an extensive discussion of the philosophy of not reviewing papers, plus an interesting quote from Professor Callaos' editorial in the first issue of the Journal of Systemics, Cybernetics and Informatics. He also introduces a new conference: the IPSI BgD multidisciplinary, interdisciplinary, and transdisciplinary conferences.

Mark Liberman has an excellent analysis of one of Prof. Callaos' papers.

Prof. Callaos comments on the bogus papers.

Look in the comments of this Blogcritics.org post for Michael Schmidt's summary of this year's WMCSCI reviewing process (note: comment authors' names are before their comments).

Greg Elin at this blog brings up one of my main beefs: that the WMCSCI folks deliberately break their spamference into a host of "sub-spamferences", thus making it harder to determine that they're all bogus. I count 11 "collocated" or "related" conferences on their web site. I'd say that I don't have the time to investigate further, but the truth is I'm just too lazy to bother.