<body><script type="text/javascript"> function setAttributeOnload(object, attribute, val) { if(window.addEventListener) { window.addEventListener('load', function(){ object[attribute] = val; }, false); } else { window.attachEvent('onload', function(){ object[attribute] = val; }); } } </script> <div id="navbar-iframe-container"></div> <script type="text/javascript" src="https://apis.google.com/js/plusone.js"></script> <script type="text/javascript"> gapi.load("gapi.iframes:gapi.iframes.style.bubble", function() { if (gapi.iframes && gapi.iframes.getContext) { gapi.iframes.getContext().openChild({ url: 'https://www.blogger.com/navbar.g?targetBlogID\x3d13597791\x26blogName\x3dAAAI-05++Blog\x26publishMode\x3dPUBLISH_MODE_BLOGSPOT\x26navbarType\x3dBLUE\x26layoutType\x3dCLASSIC\x26searchRoot\x3dhttp://aaai05blog.blogspot.com/search\x26blogLocale\x3den_US\x26v\x3d2\x26homepageUrl\x3dhttp://aaai05blog.blogspot.com/\x26vt\x3d-7311799087187756948', where: document.getElementById("navbar-iframe-container"), id: "navbar-iframe" }); } }); </script>

Monday, July 11, 2005

Minksy disappoints

I made a point of getting up early enough today to catch Marvin Minsky's keynote address. I was more than a little disappointed. He spent the first ten minutes playing with the formatting on his computer. For some reason, he prepared his talk using Microsoft Word and seemed unfamiliar with the procedures for opening documents, resizing windows, or adjusting the zoom. Then, in the middle of his talk, he paused for several minutes while trying to figure out how to prevent the screensaver from coming on. Presentation issues are common at these conferences but are usually just a minor annoyance. But when 1000 people are waiting to hear your every word, I think it shows poor preparation and a lack of respect for the audience.

The content of the talk did not redeem him. He made a lot of vague blanket criticisms about the current direction of AI such as "we need systems with better common sense reasoning" or "we need multiple, context-depedendent representations." I don't find these kind of remarks constructive or insightful. Everybody knows that these features of human behavior are critical to intelligence. The hard questions are how to recreate those features in computational systems; Minsky did not appear to have any concrete answers. He also took on the thorny issue of consciousness. I had trouble following his premise but he seemed to be saying that people like Chalmers and Searle who argue that the existence of subjective experience is an important unsolved mystery are wrong. He got a big round of applause from the audience (not surprising since Searle is so widely reviled among AI researchers) but did not actually make a cogent argument in support of his assertion, or at least not one that I could follow.

A keynote address is a grand opportunity: the lucky speaker has an opportunity to galvanize an entire community, to excite their passions, and to unite them behind a common purpose. It's too bad such an eminent figure could not make better use of it.

14 Comments:

Anonymous Anonymous said...

Disclaimer: I'm a Minsky fan and always have been. His work, character and style motivated me to get into the field.

This talk was classic Marvin, in some ways. He never was one with an agenda to sell. He says what's on his mind and I think he's earned the right to have his ideas carefully considered.

My reaction to the A/V issues was that someone involved with the logistics of the session should have helped get his laptop set up and configured beforehand.

Among of Professor Minsky's admirable characteristics have always been that he is not afraid to take on big ideas, he says what's on his mind, and he's not trying to convince the audience that he has all of the answers.

I found several ideas in this talk to be novel, intriguing and worth thinking about. That's more than I can say for many keynote talks I have heard.

Your mileage may vary.

9:50 PM  
Anonymous Mentifex said...

Did Minsky mention that AI has been solved in theory and in AI Mind software?

10:28 PM  
Anonymous Anonymous said...

No, no one has mentioned mentifex.

10:30 PM  
Anonymous Anonymous said...

25 years of the AAAI is at Robots-dot-Net.

10:48 PM  
Anonymous Anonymous said...

I agree with both the disappointment expressed in the original post and also the part of the comment saying that this talk was classic Minsky. Which makes me think: Why would the conference organizers invite him to give the keynote address when he is well-known to have poor presentation skills? A keynote address should have both solid content and solid communication. It's unfortunate that many AAAI attendees are thinking, "I might have been in the same room as a great thinker, but I didn't learn anything."

9:00 AM  
Anonymous Slawek said...

Actually, I completely disagree with the criticism.

While I agree the computer problems were a bit annoying, I have seen worse and I can understand Prof. Minsky's problems with recovering from Word crash.

I found the talk itself very interesting and very stimulating. I did not get all the points, certainly, but the general ideas he was talking about I found both interesting and novel.

Certainly I wasn't expecting him to solve the AI for us, but I expected him to point out to some problems in our approaches -- and this is exactly what he did.

All in all, I found the talk extremely nice.

11:23 AM  
Blogger Jystar said...

I'm not ready to express an opinion one way or another about the talk. Obviously, there were both high and low points.

There is, however, one issue that I would like to raise for discussion. When talking about the different levels of the brain, Minsky seemed to indicate that there were six levels. I can't remember from his talk, and I can't seem to find the slides on his website, but I don't understand where there should be six. Don Norman and others have proposed a three layer model. To me, it's not the number of layers that's the problem, but rather putting a finite bound on it. Why is it that there is a certain number of layers? How do you decide where to stop?

This is similar to the self-referential (self-symbol) looping problem. If I can represent what I think others are thinking, and I can have a representation of myself, then I can have a representation of what I'm thinking. This includes a representation of myself, which includes a representation of what I'm thinking, ad infinitum. Furthermore, if I can model what you are thinking, you must be modeling me, and what I'm thinking, etc. How does one avoid this problem? Do you simply have a recursion check, or should the self-symbol be somehow different? Hofstadter intimates in various places that it is this very self-referential property that gives people their consciousness, but I'm not certain he's very explicit as to how that happens.

4:25 PM  
Blogger Shimon Whiteson said...

I'm glad these blogs have generated such interesting discussions. I hope AAAI will do this again next year.

I'd like to hear more from the people who liked Minsky's talk about specifically what ideas they found valuable or insightful.

9:29 AM  
Blogger BSpeer said...

This comment has been removed by a blog administrator.

6:19 PM  
Anonymous Anonymous said...

I think the levels of the brain Minsky mentioned were not to give a finite ammount of levels, but rather to show that all the subfields that are developing in AI need to be incorporated in the solving of the actual intelligence problem. I wasn't able to make it to the Hawkins talk, but from what my conference buddy summarized to me, Hawkins showed that there were many different levels when it came to forming different concepts/ideas and that the ammount of levels could grow.

By the way, I'm fairly new to the field and going to my first AI conference, expecting something from the keynote, (everyone I talked to seemed to respect/adore/worship him) was quite disappointing. I'd assume the presentation would go smoothly, with understandable slides. However, Minsky did not come up to par with these basic presentation principles. It wasn't just him though, there were several papers presented with the same problems.

6:22 PM  
Blogger Arthur said...

The six levels of the
brain are the six Laminae -- Latin for "layers." Marvin Minsky founded the field of AI, and he knows whereof he speaks. Hey, did anybody see William Gibson writing in the current issue of WIRED Magazine and referring to "(so-called) cyberspace" -- a term which he himself coined in the 1984 novel Neuromancer? When you're a Marvin Minsky or a William Gibson, you can fumble around before an audience of 1,000 and the cognoscenti in the audience will NOT be disappointed. Now if your speaker is the villainous GWB, you should stand up and turn your backs to him. Over and out.

10:55 PM  
Anonymous Anonymous said...

I'm not a student, but a researcher in the field of AI and quite aware of Minsky's reputation. Reputation is not enough to overcome bad presentation skills and a lack of content in the talk. I, and several others who attended, were also hugely disappointed in this keynote speech, of which I retained almost nothing, despite straining to both hear him and comprehend his slides. I echo the original poster's disappointment.

12:42 AM  
Anonymous dustin said...

Those interested in a better understanding of Minsky's six layers of mind and his explanation of why consciousness is so difficult to think about should check out his book, The Emotion Machine.

It's due to be published early 2006, however, Minsky has made drafts available on his website: http://web.media.mit.edu/~minsky/

10:50 AM  
Anonymous Anonymous said...

i would like to comment on Shimon's remark that "everybody knows that commosense is an essential part of intelligence, we do not need Minsky to talk about this". Well, Minsky is actually PROGRAMMING what "everybody knows" and i think he is one of the very few (along with his students) to do such a thing.

4:14 AM  

Post a Comment

<< Home