AI

The hot place to be on the web this week is the AAAI-05 Blog being maintained by students at this year’s national conference on artificial intelligence. Full of deep thoughts and robot pictures and speaker controversy and bitching about the beds at the nearby college dorms. Slap a nametag on your shirt and it’s just like being there! Seriously, I am having a good time at the conference, and it’s jump starting some ideas I had floating around in my brain that got pushed to the side by classes. It’s also nice to see some old friends from grad school, … Continue reading Window on AI

Summers are all about projects – like robot building! While I understand the theory of electronics, I’ve done very little hands on, and I have a mediocre intuition for it, so this is a good learning experience for me. Nothing too ambitious going on here, we’re just following along with the instructions for building a line-following robot from Cook’s Robot Building for Beginners. Here are some photos of the prototyping process to date: To start, we have just a simple switch on the breadboard to connect the battery to the buses and light an LED when the power is on … Continue reading Robots at Home

I really ought to know better than to expect sensible commentary on computer science from Wired, but their The Other Turing Test article crossed the line of agitation for me. I’m fine with them deciding to recount a project by some undergraduates to replicate the male/female imitation game scenario in the Turing Test, and adding ALICE into the mix is interesting, though I think it misreads Turing’s intention to say that the computer has to pretend to be female, as compared to just human. But to say: “Scientists studying artificial intelligence have long argued over the meaning of this gender-bending … Continue reading Wired’s AI is Weak

And you thought learning Java was hard: Perspex is a new AI-focused language in which “it is a new, geometrical computer instruction that looks like an artificial neuron. Any existing computer program can be compiled into a network of these neurons.”. I am clearer on the use of this in practice for its potential robustness than on the argument for it with respect to AI. Put this on the shelf as something to read more about “when I have time”.

It doesn’t matter if you hate chess, you must check out this chess playing applet. The gimick? While the computer is thinking, you can see the lines representing the different moves it is considering – color coded for you and the machine. Very, very cool. Very interesting to see which moves take longer for it to figure out than others. And pretty. [via #!/usr/bin/girl]

Ooooo – the technology behind this new Chung-Kwei spam filter seems really sexy and the results sound good. Borrowing genetic sequencing techniques is a smart step beyond the Baysian filtering type tools, which the spammers are starting to subvert with those word lists at the end of messages. It’s not immediately clear to me how this algorithm avoids that problem, though. It would seem to also fall victim to messages which are substantially made up of non-spam random content.

Looking at whether representation dictates expressibility in the human mind, a study was done of the ability of a Brazilian tribe without language for distinct numbers to perform counting activities, and it was found that the ability was severely restricted. The activities took the form of creating same-sized piles or matching the number of items into a jar with the number of items removed, and they generally did not succeed if there were more than a couple of items. The news article implies that there is a lack of ability to count, which seems too broad a statement – it … Continue reading Representation and Expressibility

RobOlympics! Wooo! Sumo Robots! Fire Fighting Robots! Space Elevator Ribbon Climbing Robots! Combat Robots! It’s BattleBots meets the Mars Rover! It’s got flames, flying debris, and LEGO! It’s geek nirvana! Woooooooo!

On the one hand, this is just a story of another researcher using a web game to obtain data – in this case, the ESP Game, which anonymously pairs two users, shows them the same images, and asks them to enter descriptive words until they get a match. But there’s both some good and some bad computer science that seems to be going on here. On the good side, there’s definite value in finding ways to use people for data collection in places computers fall down, and this program seems engineered to do it well. Asking people to agree on … Continue reading Community Annotation

I’ve heard a few talks about the research in this article about analyzing text collections (such as e-mail!) for “word bursts” to organize texts or identify trends over time, and I think it’s really cool. The article even includes a speculation about applicability to weblogs for tracking social trends (and notes that Google already does something like this, though no mention of Daypop which is sort of the same though just using URL’s), but integrating this into an e-mail organizational tool intrigues me the most (which is described in a little more detail over at Scientific American).