Best. Ad. Ever.

INFO SCIENCE: The following billboard has been showing up a few places. (Seattle and Silicon Valley, from what I've read. I must say, that billboard along a highway is pretty goddamn dangerous.)







Pretty cryptic, no? Well, I'm not going to spoil it for you (unless you want me to) but I'll just say that solving it leads to another question. This puzzle is even more cryptic and I personally would never have the patience to solve it. If you would like the answer explained by someone mathier then me, click here. (If you would like me to never use a fake word like "mathier" again, please email me.)



But if you want to bypass all these profundities, cut to the chase and read the final message.



It's interesting; I used Google to solve a puzzle by Google, without actually doing any of the hard work myself. This is the future people!

Math Primers

DATA: While cleaning out old files I ran across some various old cheat-sheets and primers for some of my college math classes, mostly pre-calc stuff. They can come in handy when you want to model something like, oh... I dunno, the temperature of your house or the forthcoming economic collapse.







Definition of a Derivative



Exponential Modeling



Limits and Continuity



Linear Motion



Natural Logs and e



Parametric Equations



Sinusoidal Functions



They may be owned by one of my old TA's or some book, so don't go telling on me. ;)

Old Personal Web Site Design

INFO SCIENCE: This was an initial design for a personal webpage a few years back. I never ended up using it, but if anyone wants to snap up any of the graphics for other projects, be my guest. (Click for full view.)











I never really liked this layout style (with categories listed along the top a la Amazon), even though I've used it before and probably will again. It's a very limiting practice and most sites are better simply listing the categories in the body of the page, like Google's Directory Page. This setup gives you room to expand and also give you room to use bread crumbs (tips and examples of the content) for each category. (And bread crumbs are always a good idea.) Alternatively, you can list categories along the left column, like The New York Times, however, this can be a bit cramped and your descriptions must be terse.



The other big problem with this sort of site design is the big bounding boxes. What was I thinking? Nobody should still be using these. It's just clutter for the sake of clutter. It's also a cop-out; just make your webpage work at any resolution. Whenever I use grids now, be it html or a spreadsheet, I use either a very light grey or a medium yellow. Just enough to guide the eyes between sets of related information and for grouping groups without making the grid the focal point instead of the content or data.

Doing the Numbers

DATA: Salon is carrying an article that takes a deep statistical look at the poll numbers for the upcoming presidential election. Here David Gopoian makes a persuasive case that Kerry could be set to win the election, provided he understands which voters he can and must win over.



Most interesting for me was this part about candidataes and education level. Interestingly, Kerry has more support then Bush among people with a high-school degree or less and those with a college degree or more. Who does that leave for Bush? College drop-outs of course.

Finally, the data related to levels of education provide some interesting patterns that illustrate how the use of partisan benchmarks can shed different light on the obvious. These ABC News-Washington Post poll data show that Kerry may be en route to assembling a top-to-bottom coalition. Pluralities of both high school graduates and college graduates support Kerry. By contrast, the middle educational category, those with some college but not a college degree, supports Bush.







[...]



Beyond that, this chart demonstrates that nothing is as it would be in a typical year. Kerry leads among the least-educated voters, as a Democrat should, but is running nine points below expectations. He should be trailing among the middle group, but not by a staggering 14 percent -- which is 10 points below expectations. And neither candidate is following the script for college graduates. The Republican should be leading by 10 points there. Instead, Bush trails Kerry by four points and fares nine points worse than expectations. Kerry is running five points better than expected.



Nevertheless, these data reinforce much of the conventional wisdom. Bush's disdain for complexity and nuance is costing him the support of voters with a broader understanding of the world. Kerry's patrician manner has not yet won over voters wary of his background and style. They await nonverbal cues and verbal pledges that he cares about them.

So, the author is saying that polls reflect what we all already know: Kerry seems kinda snooty and Bush is an idiot. (I feel the need to yell: STOP THE PRESS!) I sure hope Kerry can communicate to the less-educated and the less-well-off that it's in their best interest to vote for him.

We Back: New Host

ASSEMBLEME: We've moved!



Sorry about the last 12 hours of downtime and the annoying problems with redirection and frames prior to that. Everything should work now. This should be home for the foreseeable future. Thanks for your patience!

404 In Style

OFF TOPIC: This has to be the best 404 message ever.







Be sure to read the page title for extra enjoyment.

I, Robot III

INFO SCIENCE: More A.I. musings by way of I, Robot impetus. (Worst sentence ever? Perhaps, but I'm going to stick with it.) Both via Ray Kurzweil's site.



The first, Asimov's Three Laws of Robotics unsafe?

"AI could improve unexpectedly fast once it is created," warns Eliezer Yudkowsky, Director of the Singularity Institute for Artificial Intelligence. "Computer chips already run at ten million times the serial speed of human neurons and are still getting faster… An AI can absorb hundreds or thousands of times as much computing power, where humans are limited to what they're born with. [And] an AI is a recursively self-improving pattern.



"Just as evolution creates order and structure enormously faster than accidental emergence, we may find that recursive self-improvement creates order enormously faster than evolution. If so, we may have only one chance to get this right."



Asimov's laws are not sufficient, said Michael Anissimov, writing in an article on the 3 Laws Unsafe site. "It's not so straightforward to convert a set of statements into a mind that follows or believes in those statements. Two, semantic ambiguity means that without personally understanding the reasons for the laws and the original intent, a robot might misinterpret their meaning, leading to problems. Third, Asimov's Laws ignore the possibility that a robot will acquire the ability to reprogram itself -- an inevitable eventuality if intelligent robots are created."

The second, Robots (Probably) Won't Turn Against Humanity, Experts Say in Their Defense. (Could I love that title any more? I'd have to say no.)

"The message is that they are dangerous and they will potentially have the ability to harm biological humans," said a New York University professor of computer science, Demetri Terzopoulos.







While Sony's robot dog, AIBO, has yet to cause harm to anyone, software developers like those at aibopet.com are selling downloadable programs to change AIBO's personalities, help him make different sounds, and even imitate movie characters like the villainous robots from Battlestar Galactica. If little robotic dogs can be hacked, some wonder if the human-sized robots can truly escape interference.

I, Robot II

INFO SCIENCE: New Scientist's Will Knight writes about I, Robot and Asimov Three (and later four) Laws:

Even if researchers are ever able to build robots with enough intelligence to comprehend Asimov's laws, they are unlikely to be implemented. Although they attracted some interest in the early stages of artificial intelligence research, the rules were quickly abandoned as too prescriptive and simplistic.



"Asimov's laws are about as relevant to robotics as leeches are to modern medicine," says Steve Grand, who founded the UK company Cyberlife Research and is working on developing artificial intelligence through learning. "They stem from an innocent bygone age, when people seriously thought that intelligence was something that could be 'programmed in' as a series of logical propositions."

(Read the full article here.)

Researchers: Excel Screwed My Data.

INFO SCIENCE: I don't know why the hell these people are using Excel to store genetic information, isn't there something better out there? (If not, then what are all those BioInformatics graduates doing with their time?) In any matter, they are using Excel, and Excel is doing what it's programmed to do: mess with your shizzle in annoying ways.

Excel is widely used in genetic research to process microarray data. A microarray chip detects amounts of protein produced from thousands of different genes, enabling researchers to see which particular gene is being expressed in a sample of diseased tissue, for example.



The errors are introduced because some genetic identifiers look very like dates to Excel. If the spreadsheet is not properly set up, it will convert an identifier, such as SEPT2 to a date: 2-Sep. The conversion, the researchers say, is irreversible: once the error has been introduced, the original data is gone.



In a paper published on BioMedCentral, Zeeberg et al explain that they noticed that some identifiers were being converted to non gene names.

The problem here is obvious (to anyone that doesn't work at Microsoft anyway). Excel should only reformat the cell's data if it suspects it's a date, not actually change the underlying data. This has always annoyed me about Excel. "1-1" can mean a number of things: 1 for 1, 1:1, 1-1=0; but to excel it's always "Jan-1-[this year]." If you go back and try to format the date back to being a number or text, you'll find it has been trashed and replaced with something like 37987 (which is the number of days from 1/1/1900 to 1/1/2004).



Someday -- hopefully by then end of the century -- we'll all have competent office software.

ISP Smackdown!

INFO SCIENCE: In a forgettable article in today's New York Times, reporter Saul Hansell talks about how MSN has managed to increase revenue despite adverse... zzz... Oh sorry, I was drifting off there from a complete lack of interest.







What was interesting though is how AOL and MSN are both still loosing market share. Even more interesting is that these big-four ISP's have been collectively loosing market share since their peak in '02. Where do you think the other 55% of internet connections come from? Is there really still that big of a market for mom-and-pop shops? I suspect work VPN is a big chunk of people, but not that big.