Sunday, March 25, 2007

Fact Checking

I really don't know how many people are reading this blog who don't
know me personally, but those who do know me personally probably know
that I don't have a high opinion of journalism. In this essay I
describe one of the many reasons for this.

Journalism has a method called "fact-checking." It works like this. A
article says this or that, and some fact-checkers working for the
paper or magazine or TV station or whatever make sure that what is
being said in the article is in fact true. That is, they go over all
the "facts" that the article states and makes sure they are
supported. Overall this is a great idea.

As made clear by the film "Shattered Glass," the reliance of
fact-checkers on the field notes of journalists is
problematic. However, I'd like to talk about a different problem.

I often see statements about society or individuals in magazines that
I often flat-out think are false. At any rate, I think these things
cannot be verified by a fact-checker. In fact, I doubt the
fact-checkers bother to check them at all.

For example, a few years ago I read an interesting article in The
New Yorker
about SUVs.
The jist of the article was that while SUVs
were safer for you if you got into an accident, you were more likely
to get into an accident if you drove one, because they were less
maneuverable. Fact is, statistically, a Jetta is much safer, because
you can steer around problems and avoid accidents. So far so good.

The problem was that the article said: "The S.U.V. boom represents, then, a shift in how we conceive of safety—from active to passive. It's what happens when a larger number of drivers conclude, consciously or otherwise, that the extra thirty feet that the TrailBlazer takes to come to a stop don't really matter, that the tractor-trailer will hit them anyway, and that they are better off treating accidents as inevitable rather than avoidable."

I don't believe this is true at all. I've talked to several people
about this, and though I have not done a controlled study, most people
seem to have never even considered that being in an SUV could
increase your chance of getting into an accident. Now, the author, Gladwell, admits that the conclusion might be unconscious, but this is far from obvious to me. It could be that the trade-off was never considered, consciously or unconsciously. They are acting as though they have concluded this, but that's something different. He's speculating about their internal mental state.

Saying that people place a higher priority of one value over
another is a psychological claim (and a sociological one) that could
indeed be tested. That is, it's a fact that needs to be checked. And
if it's not checked, then the journalist has no business stating it as
a fact.

The Economist is an interesting magazine because every article
is an editorial. That is, almost every article in the magazine
includes the opinions of The Economist (they have no bylines,
which one of my peevs about the magazine). On the one hand, I like
this about the magazine because it makes the articles so much more
interesting. On the other hand, if it's an editorial, which statements
get checked as facts and which do not? It might be the opinion of the
author(s), whoever they might be, that people are this way or that
way, and in an editorial, you can go ahead and state it. But as a
reader, you don't know if it's just an opinion, or a fact that has
been studied and checked.

People have a hard time imagining that psychologists can actually
empirically show psychological facts. But they can. Journalists should
not be printing statements about psychological states that are not
backed up by research, unless the article is clearly an editorial. And
even in editorials, they should try to be specific about what "facts"
are true and which are the author's speculations.

Tuesday, March 20, 2007

My Lecture On The Future of Artificial Intelligence

I was an invited panelist at a conference called "The Spotless Mind? Policy, Ethics & the Future of Human Intelligence" on February 16, 2007, at the National Press Club in Washington D.C. I was speaking generally on the next 20 years of artificial intelligence research, in terms of technical achievements, funding, and ethics.

If you wish to see my lecture, go to this website
and watch part three. My lecture is first and lasts 15 minutes.

There should also be a book coming out based on the presentations at this conference.

Sunday, March 18, 2007

Repetition / Meaningfulness Tradeoff

My friend Daniel Saunders and I have a minor disagreement on how much
novelty a person should have in his or her life. It's only a minor
disagreement; we're reasonable people, but I will make it sound like
the both of us are more extreme in our views for purposes of this

Daniel says a person should maximize his or her novel
experiences. There are several good reasons to do this. First, novel
experiences expand you mind. They give you new ways to look at things
and make you a more interesting person.

Second, novelty likely makes your life seem longer. One of the popular
theories of how we perceive time passing is through change- if we do
the same exact thing every day, we don't remember every one of those
days individually, we compress it in our memory and just sort of
remember it once. This might be one part of why time seems to go by
more slowly as we age: Our lives grow more routine, and we've seen so
much that fewer things seem new.

Daniel lives by this philosophy. Even his favorite movies he's seen
only a couple of times.

I say that repetition is the only thing that makes life
meaningful. Imagine how empty your musical experience would be if you
never heard the same song twice. Think of how many albums you didn't
like at first but became your favorite only after repeated
listenings. Sometimes you need to experience new kinds of things
multiple times to get to appreciate them- cuisines, hairstyles, new
art forms.

Further, as you re-experience things they become deeper and more
meaningful to you. That is why rituals and traditions are so
comforting. Think of a holiday that means a lot to you-- Halloween is
coming up-- and imagine how much less meaningful it would be if you
did completely different things every year.

This disagreement has made me think about how I want my life to be.
How much novelty should I strive for in my life?

When it comes to movies, Daniel and I definitely differ. For one
thing, he seems to have a much better memory of what he's seen. Years
later he knows character names, and lots of details that are totally
lost to me after a couple of weeks or months. For me, more than
Daniel, movies are like music. They need to be seen a few times to be
really appreciated. I've seen my favorite film (Kiki's Delivery
Service) about six times, often with different people I want to show
it to. I've seen Strange Brew more than any other movie, I
think, and I look forward to seeing it many more times before I die.

When I see too much novelty I tired out and stop enjoying myself. I
see a lot of theater, but one of the things I don't like about it is
that it's so ephemeral-- once a show closes you can't see it
again. When I see a show I really love, I often go see it again and
again. If a script is what makes the show good, then you can often go
see a different production, but if the actors or the direction or set
is what makes it, you're stuck with just your memories for the rest of
your life. This is a reason, in my mind, to give movies priority over
theater. If you love a movie, you can see it again and again and
enrich your and your friend's life with it forever. I saw a wonderful
show in Edinburgh called Roadmetal Sweetbread. It featured two
actors interacting with each other and their own video projections. It
was magical, wonderful, and I'd love to see it again. As it is, month
by month I remember that it was great, but I am forgetting why.

I don't think there is a resolution to this problem that will work for
everyone. Nor do I think it's the same for every kind of
experience. For most people, it's important to hear music multiple
times to maximize what you get out of it. For everybody, as far as I
know, never talking to the same people more than once would be an
extraordinarily empty life.

But for movies, travel destinations, breakfast menu selections,
holiday activities and lots of other things, it's a gray area. There
is a trade-off between exciting new stimuli and the meaningfulness of
our experiences. We all should consider the problem and make a choice
about how much novelty to seek in the different aspects of our lives.

See also my post on repetition in art:

Friday, March 16, 2007

Bornean clouded leopard: new species of great cat

I don't normally post news items here, but I've been so excited about this newly demarcated species of great cat, the Bornean clouded leopard! It's so beautiful. It's not newly discovered, but has recently been determined to be a species all its own.

See the news article at:,,2034219,00.html

Tuesday, March 13, 2007

Missing Poetry

I've come up with a new kind of poetry called "missing poetry."
Here are the rules:

1. Each line is a commonly-known phrase or expression
2. One word is missing from each phrase or expression, leaving a blank
3. This blank space is down the center.
4. The poem is, in some sense, made of the missing words down the
middle, which the reader infers. These are the "missing words" in the
overall "missing poem."

For example:

chance favours prepared mind.
take me to the , drop me in the water.
religion is the opiate the masses.
what's but a second-hand emotion?
haste waste.
what's name?
I pray to God my to keep.

The missing words are "the river of love makes my soul." Note that
"what's my name?" Could also be interpreted as "What's your name?" in
which case the meaning of the poem is a little different. In this way
the poet can take advantage of the ambiguity and have the reader make
it a bit more his or her own.

Take line two, which is a lyric from the Talking Head's song "Take Me
To The River." Even if the reader didn't know this song, he or she
could guess as to what was in there. She might guess "Uh, ocean?" and
that wouldn't be so bad either, as a missing word. A good poet can
anticipate what a reader might put in the blank if the phrase is not
recognized. Poetry has a lot of reader interpretation regarding what
the meaning of the words are; missing poetry has interpretation
regarding what the words themselves are!

Each missing poem is culture-specific. Someone who spoke English but
was not steeped in my culture might not know the "correct" missing
words, but that's okay. It's okay to make culture-specific art (like a
series of paintings
based on the videogame Pac-Man
), and it's also okay for the words
to be "wrong." Thinking of which words might fit, and finding the most
meaningful missing words, for you, could be fun.

It would be really great to have each line have some thematic
relationship with the poem's meaning.

Now, I see an AI application here!

First version: You get a database of common phrases. The user inputs
the intended missing words, and the AI outputs the missing poem. The
AI would suggest phrases for the author to use. In the above missing
poem, I originally wanted "makes" to be "shapes," but I could not
recollect a phrase with the word "shapes" in it. A program could help
with this.

Second version: Using some kind of semantic distance, the AI chooses
the best phrases based on the meaning of the missing words.

Third version: The user inputs a theme (in the form of some set of
words) for the missing poem, that might be even at odds with the
missing words, for irony or whatever.

If anyone would like to work with me and program this, let me
know. It's simple enough that it could be done over email. It involves
finding the database of phrases, incorporating some semantic distance
measure (we can probably find one on the web, use LSA, or WordNet or
something), and doing some not-too-difficult coding. If you want to
work on this, send me an email at

Everyone else, feel free to add your own missing poem as a comment.

a wrap.
Give me your lovin', all your hugs and kisses too.
different strokes for different !

Sunday, March 11, 2007

Motivating Examples

It’s always good to have good motivating examples. I was in a talk once at the case-based reasoning conference, and this guy’s example was an internet search mechanism for finding information about cars. The search took like 13.4 seconds or something. He said that was too long.

It is? I might do a couple of these kinds of searches every four years, and 13.4 seconds is too long? He spent the next 30 minutes telling about how he shaved it down to like 2.6 seconds. I spent much more time watching this part of the talk than his findings would ever save me in the real world.

Luckily for me I some of that time writing this essay, during that part of the talk, so it wasn’t a total wash.

Perhaps his methods apply to a more interesting, larger class of problems, but damn, he should have chosen a better guiding example; because the stupidity of the whole thing distracted me throughout the whole talk. I had to keep reminding myself of his numerous accomplishments to keep me from thinking he’s a total dumbbell.

Friday, March 09, 2007

Amazing Book: Codex Seraphinianus

I think I have found the strangest book ever published. It's called CODEX SERAPHINIANUS by LUIGI SERAFINI. It appears to be an illustrated description of another world, written in an unknown language. I find it very inspiring, a work of mad genius. Please see some pictures from the book on the web:

I heard about it in Hofstadter's Metamagical Themas, which was great.

The Codex might be available in large libraries. Carleton had a copy. It's a very rare book, just about impossible to buy, oversized, printed on beautiful paper, and full of color illustrations. If you have the heart of an artist, I highly recommend you check it out.

Sunday, March 04, 2007

Some of my scripts will be performed in "Deviants"

Several of my short plays are being produced in Kingston, Ontario March 6-10. If you're in the area, check out "The Deviants."

Why I love Sampling

One of the cornerstones of hip-hop music is sampling. I've talked to
many people over the years who find sampling uncreative and
unmusical. Personally, I think sampling is a thrilling new kind of art
form, and I hope to convey some of my enthusiasm in this essay.

A critic of sampling says that the sampling artist does not create his
or her own music, but rather takes good ideas from other artists and
re-packages them in a very slightly modified form, and passes this off
as his or her own work. And, I have to say, there are examples of this
where this criticism is dead on. Let's take M.C. Hammer's track "U
Can't Touch This." In this track, the bassline is taken from Rick
James's classic "Superfreak." The problem is, one of the things that
made "Superfreak" so wonderful was, you guessed it, the bassline. So
taking what made the original work wonderful and using it to make your
own work wonderful is not being particularly creative. "U Can't Touch
This" is a good track, and better than many tracks out there, but it's
good in the same way that a print of the Mona Lisa is good, and will
probably look better than most original paintings you could afford to
buy. For Hammer and the print of the masterpiece, the genius of the
original comes through in the reproduction.

Sampling, at its best, is when the sampling artist finds something
that is perhaps unappreciated in the old work, and turns it into
something wonderful and new in the new work. A good example of this is
the drum loop called "Funky Drummer." The Funky Drummer is taken from
a James Brown track. Why is this not the same as the Superfreak
example above? Because 1) the sampled drum sequence is a minor part of
the original track, 2) it doesn't loop in the original but does in the
new, and 3) it's slowed down. You end up with something catchy, new,
and just about unrecognizable. The Funky Drummer takes something
groovy and small, and turned it into a fabulous beast in its own

Examples of sampling exist at every point between these two
extremes. So to me, it's unfair to criticize sampling in general as
being uncreative. It can be uncreative, but, god knows, so can
some jerk on a guitar.

Sampling came onto the musical scene so quickly that the law had to
scramble a bit to catch up with it. It did, though, and now all
samples have to be cleared with the original artist or label to be
released on an album. This is good, and fair, but is not without its
cost. Certain bands like ABBA just about never release sample rights
to anybody, which, in my opinion, is holding back on all the possible
creativity that could have resulted from sampling ABBA. (They made an
exception for Madonna's "Hung Up," simply because the track was so

Anyway, there was a time when people sampled whomever and there were
pretty much no repurcussions. And during this time an album came out
that took sampling to another level, and created, in my opinion, the
greatest sampling work of art ever made: "Paul's Boutique" by Beastie
Boys. Paul's Boutique is so heavy in samples that people are still
trying to track them all down
. There are webpages dedicated to the
task. During a time when rap was a bit pared-down and minimal, this
incredible album dropped and just blew people away. Anyway, it's
generally acknowledged that "Paul's Boutique" would be impossibly
expensive to create today, given the enforcement of sampling
laws. It's a relic from an era that's gone forever. Hip Hop today is,
by financial necessity, less sample-based than it was. So in terms of
sampling artistry, "Paul's Boutique" may never be topped. Sigh.

For those of you who are still not convinced that sampling is a
legitimate art form, I want to appeal to two other kinds of sampling
from the visual art world that you might appreciate: collage and

Collage is not typically the kind of thing that people get all excited
about, except for the stuff Terry Gilliam did for Monty Python's
Flying Circus. I challenge anyone to watch his stop-action animated
collages and deem it non-creative.

Photophaphy is an interesting case of visually "sampling" from our
environment. Like a bad music sampler, a photographer can "cheat" and
have an interesting photograph of something that is widely acknowledge
to be visually interesting-- a photo of the Mona Lisa, for
example. The best photography, though, chooses as its subjects those
parts of the world we never would have looked at before and says "Look
at this, see how beautiful or interesting it can be." I love it when
I see a gorgeous photograph that shows off the artistic eye of the
photographer-- she saw a lovely composition in cracks in ice, or in
candlewax drippings, or in the way the light played on an old man's

Musical sampling, at its best, is good for the same reasons. So open
your mind to it. There are lots of fascinating examples of creative
use of sampling out there to be appreciated.

And check out "Paul's Boutique" (Which I think is one of the top 5 albums ever).
If my admiration of the album does
not impress you, know that no less than Miles Davis said he never
got tired of Listening to it.