A good proof is like a story. Both are like reality run backwards: disparate things flow to an ordered final state.

Take-the-Best

This is continuing the ideas in a previous post, Franklin's Perceptron. I am trying to find out whether any ancient methods of artificial intelligence actually worked. Any ancient methods of AI must have been very simple, but a surprising finding in machine learning has been that simple models can often perform better than more complex models.
One example of this is the rule called take-the-best. Here's how it works:
Suppose you want to predict whether one team or another will win a soccer game. (Maybe you're going to place a bet.) You have various pieces of information about the teams: the ages of the players, the number of wins each team has had this season, and so forth.  You also have this same kind of data from the whole league from a few seasons ago, along with which teams won and which teams lost. We call this the training data.
The take-the-best rule goes like this:
See how well each piece of information does at predicting wins on the training data. For example, the team with the most past victories won 75% of the time.  The team with the oldest players won 64% of the time (say this is a youth soccer league) and so forth.
Now, to make your prediction, just choose the piece of information that does the best job at predicting wins, and go with that.  Ignore all the rest of the data. (If that piece of information says they are both as likely to win, you can use the next best predictor as a tiebreaker.)
Now it's easy to see that this is a good quick-and-dirty method that should work pretty well.  The surprising thing is that in tests with real data, this method usually does better than statistical regression, artificial neural networks, and all kinds of other complicated techniques people have come up with to solve this kind of problem.  It's as if the food they served on the dollar menu at McDonald's was not just the cheapest food, it was also the tastiest and the most healthy!
 The reason it works so well seems to be that it avoids overfitting.  It's easy for more complex methods to find patterns where there are none, and be misled by noise.  Here's a paper that explains more.
What I'm looking for now is people who advocated this method, the longer ago the better. Even better would be if they built some kind of device to do this for them. And if not, it's at least would make a good basis for a science fiction story set in an alternate past.

The First Search Engine

(Excerpt from Machinamenta)

The idea of punched cards for storing data was taken up by a Russian inventor, Semyon Korsakov, in 1832. Korsakov worked in the police ministry and kept extensive records on the populace. He devised a way of recording data by means of holes punched in cards. If two cards shared many of the same holes, it meant that the data they contained was similar. This fact could be used for creating a kind of search engine, where a plate with pins would rapidly scan across hundreds of records, only falling in and stopping when the pins could fall through all of the holes, indicating an exact match. Korsakov was excited about his ideas, and thought that they could be used to enhance human intelligence in the same way that the microscope and telescope had been used to enhance human sight. He wrote,


"machines intellectuelles would limitlessly strengthen the power of our thought, as soon as distinguished scientists apply their knowledge to studying the principles of this process and compose the tables necessary for its application in various fields of human knowledge." 

His designs for the ideoscope and homeoscope were released to the world (open source fashion) rather than patenting them to encourage their widespread use and further development. Unfortunately the Russian Academy of Science didn’t see the potential and little resulted from the inventions. He is today better remembered by the homeopathic medicine community for his remedies, than by the information science community for his ingenious method of searching through the database of those remedies.

For more about him, check out the Wikipedia article (and also the references at the end of the article):
http://en.wikipedia.org/wiki/Semen_Korsakov

I like this because it is basically the original version of Google. We don't think of search engines as AI, usually, but they really are automating, in a spectacular way, something that was only possible before through human intellectual effort. Like topics in philosophy, once something is solved we no longer call it artificial intelligence.

Film and Photography as Art

I was reading an article by Berys Gaut on Cinema as Art.What brought me to this book is the possibility of a naturally occuring camera.  The author imagines a natural camera obscura (a hole into a cave) and some naturally occuring photographic salts on the floor of the cave. If a photograph is created of a passing elephant, we would call that a real photograph. But if paint were to fall onto the floor because it was blown by the wind, we wouldn't call that a real painting. That's because a painting, Gaut supposes, necessarily involves intentionality on the part of the artist, while the photograph does not.
Could one create artificial intentionality in a robot that puts paint to canvas?
He points out that in a sense, any film is a documentary, in the sense that it records what the actors were doing at certain moments in time. Except special effects blur this line, as do animations, I would think.
Anyway, there may be a much older connection between artistic representation and the camera.  Watch this video about stone age cameras!

How can we compose better music?

I just read this excellent article about the causes of musical frission, what makes a particular moment in a piece of music strike you as just right. It would be nice to see these ideas built into an automatic composition program like Melody Generator. In the book I mentioned how a sunset can induce strong feelings through an unguided process.  I don't see any reason we couldn't make a system that generates beautiful music like the atmosphere generates beautiful sunsets.

In fact, it might be cool just to make an artificial sunset generator using a fluid simulation and the physics of light passing through clouds.  This is an image I made years ago using POV-ray and Perlin noise functions:

 I think it gets the cloud shapes about right, though the lighting needs a lot of work.

Physiological and Musico-Acoustic Correlates of the Chill Response

Martin Guhn, Alfons Hamm and Marcel Zentner

Music Perception: An Interdisciplinary Journal
Vol. 24, No. 5 (June 2007), pp. 473-484 

Anima in Anime

Anima in Anime: the nature of the soul in Japanese animation.

I just thought that would be a fun essay to read. It would discuss Ghost in the Shell, The Spirits Within, Paprika, maybe... what else do you think would be good?
I read a book about Ki (the Japanese symbol for breath and spirit) when I was trying to learn Kanji while living in Nagoya. Since my English reading material at the time was so limited, I have an almost supernatural memory for anything I read at the time.
Edit: I've just finished watching the anime series Mushishi. In these stories, the spirits are more like insects or fungus. They are a natural phenomenon that sometimes affect people when their lives intersect. I highly recommend it. 

19

This looks like an interesting resource on 19th century views of the body as machine.
http://www.19.bbk.ac.uk/index.php/19/issue/view/70