Archive for the 'Science' Category

Cyc and Google

In preparing for my ESSLLI course and evening lecture in two weeks, I’ve been trying to find out what state Cyc is in these days. I still can’t find any reasonable kind of technical documentation (if you know of any, let me know), but I did find a really cool talk that Doug Lenat, the founder of Cyc, gave at Google last year. I think everyone who works on knowledge-based AI should find an hour to watch this talk really soon. It has a fantastic motivating introduction on why we really need formalized commonsense knowledge, and shows a couple of things that Cyc can do. Of course, this doesn’t answer my original question, but perhaps I can use some of it for my talks anyway.

Now, the other thing that really impressed me is how easy it is to watch Google videos. I knew before that Google records the public talks that people give there and makes them available, but hadn’t really looked at any of them. But what’s really nifty is that you can use the Google Video Player (available for Windows and Mac) and have the video automatically downloaded for offline use, in much better quality than the online video. Now if I could just find a way to read just the subtitles (the entire video is subtitled) without watching the video in realtime, that would be even cooler.

Edit: I finally figured out how to access FACTory, an online game that asks users questions and uses their answers to add more stuff to the Cyc database. This seems like almost exactly the right thing to do, and gives you an idea of what sort of things Cyc knows about. You can find it here (your browser needs to support Java applets), and it’s really fun to play for a bit.


Social Democracy in Unexpected Places

Columbia University is one of the places in the world with the highest concentration of Nobel Laureates. It is second in Wikipedia’s list of Nobel laureates by affiliation, and just recently in 2006 Columbia faculty won two Nobel prizes (for Economics and Literature).

I would be crazy not to make use of this as much as I can. And so of course I went to a public discussion on “An emergent India: Prospects and Problems” between Prabhat Patnaik, apparently a renowned Indian econonist, and Joseph Stiglitz (Nobel Prize 2001), whom even I had heard of. It was interesting! They were both entertaining and interesting speakers (unlike the moderator, who only mumbled and had to be constantly reminded to speak into the microphone). Of course everybody was duly impressed by India’s GDP growth rate (although Patnaik didn’t completely trust the figures) and had a lot of economic insights to share, but what I did not expect was this outburst of unmitigated social democratic opinions from both speakers. They criticised the growing disparity between the living standards of the reasonably well-educated urban population (about 250 million people) who benefit from the rapid growth of the IT industry and the unskilled rural population (about 800 million), where literacy among adult women hovers around 50%, 35% of the population live off $1 or less per day, and 47% of children below the age of five are malnourished. They applauded China for preventing an imbalance of the same magnitude by maintaining high prices for domestic agricultural products, which they do by manipulating currency exchange rates in order to avoid introducing import tariffs and pissing off the WTO (clever people, the Chinese). And they openly advocated raising taxes on the use of environmental resources and imported luxury goods in order to finance improved education at all levels. (And Stiglitz looked positively sly when he said that the WTO doesn’t want you to tax luxury imports, but “we’ve found ways around this in other cases too”. :)

Having lived in Western industralised countries all my life, my primary exposure to the economy of India so far has been whenever they closed down factories over here and moved the jobs to Asia. I’ve always felt that this is a strategy that might eventually backfire; there is a lot to be said for the excellent infrastructure and social stability that European countries offer. But I had no idea what “social stability” really means before they mentioned today that there has been a wave of suicides among Indian peasants, who live less than an hour away from the major IT centres, and peasant riots in rural China are frequent. And it’s an interesting perspective that if the movement of jobs from Europe and the US to India is because European and Indian workers compete with each other for lower labour costs, then one would expect that the wages in India would go up, but they don’t. Instead, they analysed the current trend towards lower wages and the “renegotiation of the social contract” as an effect of capital being able to move, while labour cannot (and hence the negotiation position of the capital anywhere improves). So: Labourers of all countries, unite! Who would have thought I’d have to go to New York to hear this?

Free food

I’ve been sort of stressed in the past days because I want to submit a paper to ACL. The deadline is Tuesday afternoon next week, and until two days ago, I did not have an implementation of the system that the paper is about.

However, today was a pretty good day. Part of this is that I’ve made good progress with the writing. But on top of that, just when I was kind of getting bored and unhappy with the paper and also getting hungry, I discovered some leftovers from a project meeting in the pretentiously named “lounge” of our floor:

Img 3306 450X338.Shkl

Isn’t it pretty? See, one of the things that I really like about American universities as opposed to German ones is that free food is readily available, and although I did not undergo the rigorous training that makes American-schooled graduate students extra sensitive to free food, I’m not going to pass up an opportunity like this if it comes up and licks my face. Yay for Columbia. (However, there have been rather frequent reports recently about mice settling down in the CS offices here, and I can only begin to imagine how a mouse must react to a table full of yummy food like this …)

Life is catching up with me

Just about two weeks ago, I had the feeling that I had so much time that really the only thing that could fill it was to write this blog.

Now, life has been catching up with me on all fronts, and I find myself falling back into bad blogging habits. The pace of my work has picked up considerably, and next to that, there’s so much bloggable fun stuff going on that I hardly have the time to write about it.

So let’s just summarise the highlights of the past few days:

  • On Sunday I went to see the NextFest as planned. It was pretty cool: They had lots of robots — including one that looked so lifelike with its artificial skin and subtle facial expressions that I mistook it for a real human for a second –, and future space suits, and a device that will highlight the veins under your skin, and lots of other stuff. The singing rabbit robots were a bit of a disappointment, as they did have a darkened room with a few hundred of them, and they were supposed to perform a specially composed opera, but they didn’t sing. There were also two different video games in which the player was filmed by a camera and inserted into the game world, where they had to touch falling balls, or fight against evil sprites. This would probably not be so difficult to implement yourself, and perhaps it would be a fun little project to do in my copious spare time.The lifelike android (who reminds me a bit of Lex Luthor in Smallville; I wonder if that’s intentional):


    One of the camera-based video games: You can see the player (shown to the right) very faintly at the center of the screen to the left; watch for the colour of his shirt.


    The coolest game of the entire exhibit was Brainball, in which each player wears a bunch of EEG electrodes on their head, and there’s a little ball on the table that moves towards the player with the higher stress level. The really nice part about this game is that the little ball on the table actually physically moves, and so it makes the game look like a match between two telepaths.

  • On Monday night, we went to Rock ‘n Roll Karaoke at Arlene’s Grocery in the Lower East Side; I must say that I’m beginning to like the LES better and better for its quirky little shops and bar and cafe scene. It was fantastic: They had a real live band that would accompany karaoke singers, with admissible songs ranging from the Beatles to Iron Maiden. Most of the singers were actually pretty good, too; but the atmosphere was so friendly that even the singers who really sucked got a warm applause. It was great fun, and I’ll need to go there again and perhaps sing too. :)
  • Yesterday I went to visit Rutgers University to see a talk (a classical semantics talk with a handout, on a topic that my advisor worked on thirty years ago, but very accessible and fun nonetheless) and meet some people. I must say that one thing I really like about American universities is that they give you free food whenever they can get away with it. In this case, pretty good Thai food before the semantics talk.

Phew. Now today I’ll go run some errands, try to pick up the computer for my office, finish my slides for the talk tomorrow, and start putting redundant constraints into the constraint-based planner that I’ve also implemented over the weekend. Then tonight I’ll go to my third Big Apple Chorus rehearsal and presumably become a regular member. No rest for the enthusiastic.

If we understand it, it’s not intelligence

I’ve recently discovered the papers of Drew McDermott, one of the great old AI researchers, who happens to be one of two people in a 500 km radius who is interested in automated planning and therefore a hypothetical, potential collaborator. Next to technical papers, he has written some very opinionated and interesting papers on AI as a whole, including such famous papers as “Artificial Intelligence Meets Natural Stupidity” (written just after he got his PhD, no less).

An interesting piece on Deep Blue got me thinking. In this article, McDermott refutes the common argument that Deep Blue is not intelligent because it “only searches through millions of possible move sequences ‘blindly.'” Perhaps the real point is that because some people understand, and many have a general idea of, what Deep Blue actually does, it can’t possibly be intelligent: Intelligence is some kind of mysterious thing that happens in our heads while we’re not looking. Once we know how it works, it just becomes another man-made algorithm; the mystery is gone, hence there is no intelligence. In fact, one reason why neural nets seem so appealing as a model of intelligent systems might just be that they function as black boxes, whose inner workings nobody except the learning algorithm that trains them really understands, and therefore still retain the potential for being “intelligent”.

It’s a bit reminiscent of Richard Feynman’s definition of a trivial theorem in mathematics as “one that has already been proved”. Once you’ve figured out how to do it, it seems easy, and everybody could have done it.


May 2018
« Feb    

RSS Unknown Feed

  • An error has occurred; the feed is probably down. Try again later.