World Championships in Robot Soccer

fACTOR writes "The Robot World Cup is an initiative to encourage research in artificial intelligence and robotics by applying the new technology to the world's most popular sport - soccer. If this idea takes off, maybe pro sports salaries will drop, and there will be a new kind of job created: "sports robot programmer."

Comments

AI sucks

[Sanity]

I am not really bitter, there are many things I like about the Eduni AI dept, and about the people in it. However if there is one thing that I have learned in discussions on SlashDot, it is not to understate your case (see new subject ;-).

Further, I do not claim that some things that have come out of AI research have proved useful, linked lists, and expert systems are two good examples. But what do linked lists have to do with making something think?! While in the past there were some good ideas in AI (GAs, Neural Nets), these were done to death in the 70s and early 80s. There is nothing I am aware of now that is significantly in advance of what people were doing 25 years ago.

WRT what I know about that joke project, my knowledge is gleaned from an hour long lecture given to us by the girl that did it, she also answered some questions that I put to her as a result of my disblief at how lacking in creativity the whole thing was (although I was more polite at the time!). I stand by my belief that NOTHING was learned about intelligence from that, and many other such projects, other than how not to achieve it.

The definition of Artificial Intelligence you point to is cyclical. It defines itself in terms of "Intelligence" (see first line!). Any dictionary that used part of the phrase it was defining in the definition would be laughed at, I think this definition deserves the same treatment. Minskey's definition is the same. Show me a definition of AI that doesn't used words that themselves require just as much clarification as "Artificial Intelligence" does, and I will eat my hat (or would if I had one).

I have no problem with young fields per se, I just have a problem with young fields that aren't even trying to grow up.

I know that there are other undergraduate courses in Europe that cover AI, but how many of them are taught by an actual AI department?

PS. Where did you do your MSC? Edinburgh?

--

Re:Benefits to AI research dubious at best

[r]

all right, let's start a holy war. :)

first of all, artificial intelligence is not a theoretical field, but a science. unlike in theory fields (math, theory of cs), you can't just expect everything to be neat and clear and derivable from first principles. just remember how many centuries (millenia?) it took to come up with a reasonable model of the atom. we can't expect a reasonable model of the mind to just pop up overnight. for every brilliant insight there is a dozen detours. such is the way of science.

and secondly, re your comment of ai being an ill-defined field - it's not the field that's ill-defined, it's the definition of intelligence that keeps changing on us! back in the 50s, when first ai systems were born, people actually considered intelligence to be equivalent to formal inference, spatial reasonoing, and so on. but as computers started getting good at those, the definition kept changing, as if to exclude what computers were doing - people started realizing: what about emotions, what about social skills, what about pragmatics? but this is a vicious epistemological circle - ai trying to model intelligence which is constantly being redefined because of ai's successes. to blame ai for this circle would be as foolish as blaming mathematics for people's fear of differential equations. the question should be how to break it.

New hooligan songs needed...

[rde]

Obviously one can't go to a soccer match without some appropriate songs; some (ole ole ole) might be appropriate for a computer, others (Lq46, you're a wanker, you're a wanker) less so. Perhaps a few new ones could be written. My suggestions:

You're gonna get your fucking bolts unscrewed

You'll never roll alone

There's only 1 Kryten 523C, but he's got a 30-day guarantee and can be returned in the original packaging for a replacement

Benefits to AI research dubious at best

[Sanity]

During my degree in Artificial Intelligence we participated in a similar Lego based game, this time robot-rugby. We were given kits (not Mindstorms, but an in-house concoction) and told to build robots that played this game. During the course of the project my team and I became somewhat disillusioned with the whole thing, and while some came up with hopelessly elaborate software, we made our robot bumble about randomly, grasping the ball if detected. While we didn't win, we did come second in our section.

In my opinion most current research in AI is non-creative rubbish. I am not unqualified to make such a statement given that I went through probably the only, and almost definitely the best respected undergraduate degree involving AI in Europe, and am an ex-president of the Edinburgh University AI Society. If there was one thing I learned from it is that most people doing AI research are either kids who just think it sounds cool, but don't have the intelligence or creativity to progress the field one little bit, or they would rather talk about AI than actually do anything about it.

--

Re:Benefits to AI research dubious at best

[iapetus]

You seem remarkably bitter about the whole affair. There is some genuinely useful and interesting research coming out of Edinburgh (and out of AI departments elsewhere). AI projects can have real-world applications which affect 'grown-up' subjects such as computer science (see the work done in Edinburgh on programs such as CYNTHIA, for example).

How much do you actually know about the machine humour research, by the way? Have you read the relevant theses (info available at this page)? FWIW, 99% of the jokes the research was based on were crap, too. :)

As for your claim that "Artificial intelligence lacks a proper definition", I am slightly confused by this. There are any number of good definitions of Artificial Intelligence. They may not all agree on the exact boundaries of the discipline, but nor do definitions of many accepted fields. A definition of AI as taught at Edinburgh can be found here. Minsky provides an often-quoted definition which takes a more practical approach: "the science of making machines do things that would require intelligence if done by men". Any good AI text book (and many bad ones) spend a chapter or so looking at definitions of the field.

For what it's worth, if every field 'without foundation' were ignored, the grown-up subjects of which you are so proud wouldn't exist. What's wrong with working in a young field? That's where all the discoveries are still waiting to be made...

I feel as qualified to comment as you, given that I went through the best respected MSc course involving AI in Europe (actually there are many other courses, particularly at the undergraduate level that cover AI). Evidently I got more out of my time at the department than you did.

but why?

[cpt kangarooski]

I mean, we've already made great advances in the field of robot boxing with Rockem Sockem Robots. We should finish that before going on to make robot soccer players, and the inevitable artificial riots.

But could they be programmed to...

[nstrug]

• Shag Spice Girls?
• Gorge themselves on kebabs?
• Cry like little girls for getting yellow carded?
• Drop like a felled tree and lie writhing on the deck whenever another robot looks at them in a nasty way?
• Show enormous potential and skill before snorting half their salary up their noses and drinking the rest?

I just don't think they could compete...

Nick

Re:But could they be programmed to...

[SEWilco]

• Roll up in a ball the same size as the soccer ball and bump the real ball around. Do you suppose the opposing team might get a little confused? Might the opposing team acquire a few fouls for kicking opposing team members?