Emotional Talk

I attended AAAI last month and was on vacation so that is my excuse for zero posts last month.  A few of my posts in August will discuss the conference.  This is the first of them and it is a discussion of Marvin Minsky's keynote address.

There were two main themes of his address one was a discussion of his upcoming book, The Emotion Machine, hence the title of this blog entry.  You can find the first eight chapters of the new book by going to his main web site.  He also recommends reading Push Singh's Ph.D. thesis which can be found on Push's web page here.  I will provide some commentary on both of these over the next few months, but wanted to provide you with the references immediately,

The second theme, related to the first was that AI needs to be refocused.  He said that recently AI has gotten Physics Envy, trying to find general methods, but yet most of the programmtic successses in AI, according to Minsky, have more to do with Moore's law (computers became more powerful and could search more) than with advances in AI.   The talk was great, for example, he declared that in AI and psychology Occam's razor is the wrong idea; you are describing a complex system and it is not simple!
To quote from chapter 1 of The Emotion Machine:

So this book will embark on the opposite quest: to find more complex ways to
depict mental events that seem simple at first!

As I said, more discussion on Minsky's ideas will be forthcoming as I read the book.  Just wanted to provide you with more possibilities  to read as you hopefully take to the beaches in August.  Check back, as the Fall term starts getting together I should be posting more frequently -- and catch up posting some of my student's blog entries.   Later!

Once upon a time ...

It is easy to be cynical in this age of outsourcing, impossible schedules and managers who think the only valid motivational technique is fear.  It is hard to remember the sense of wonder of software and computers that initially lured us into this field.  In working on Research for another task I came across an article by Allen Newell, "Fairy Tales," AI Magazine, WInter 1992. Note that you have to find the article in the table of contents, the first click on the title provides the abstract and clicking on the title within the abstract provides the article. It is a short, 3 page article and I highly recommend it to give yourself a boost and get back to viewing the big picture of why you do what you do.

The article discusses computer science and the promise it provides.  Rather than summarize it, I urge you to read it.

A bit on Allen Newell.  Allen, along with Herb Simon, Marvin Minsky and John McCarthy are known as the Fathers of AI in the United States. He was a great guy and had a great sense of humor.   He also was very direct and got right to the point.   For my Human Computer Interaction courses I have a clip of him struggling with a photocopier.  His personalty shows through perfectly. He died over a decade ago but the ripples he started will continue for a long time.

To close this off I would like to quote the last sentence of his article:

Finally, I wish to express my feeling of childlike wonder that my time to be awake on this earth has placed me in the middle of this particular fairy story.

Later.

Can Open Source become too popular?

In his Log Book entry Lucas Vickers discusses his experience with Open Source support and contrasts it with his experience with Microsoft support. He also poses an interesting question at the end, thqat as more folks use Open Source the expertise level will go down and more Open Source developers will be bombarded with newbie questions. Do you think this will happen? If so, how can Open Source developers handle the support demand?

As an added bonus in his Log Book, Lucas gives us a pointer to another rule base language, Drools. In a later blog entry I will give you my impressions of the Drools tool. Later!

Lucas Vicker's Log Book Entry:

Open source is the topic of this week’s discussion, and it is something that I am fairly interested in. I have worked with open source products, and I have also worked with Microsoft products. I know using Microsoft as the model for what comes out of corporate development is a little unfair; however I am going to do so because I have the most extensive experience with them. To be honest I have found Microsoft products that work really well, and I have found open source products that work really well. I once developed a program for the pocketPC platform using Microsoft tools, and oh my that was one of the buggiest platforms I have ever worked with (this was back in 2002). I have also recently tried to integrate Drools, an open source and java based rules engine. After a few weeks of heavy coding with Drools I found that there was a major flaw in the implementation which halted my work. What I am trying to say is that each product has its flaws. The difference is as soon as I had a problem with drools, I sent off an email to the developer, met him on IRC on #drools, and we resolved the problem with in a few days. Any problem I had with Microsoft all I could do was search MSDN and pray that they have an answer to my question. I think that one major factor to consider is the group of developers behind a project. Open Source developers really love what they do, work in their spare time because it’s a passion, and are very easy to contact. Microsoft developers, however, hide behind the veil that is Microsoft and it is near impossible to get an explanation for why a function works a certain way. I think knowing that OS developers are developing because they want to and love to, and are accessible, gives open source software an edge. I wonder if this edge would disappear once the average user began attempting to use OS software and the developers became flooded with repetitive and useless questions.

Invasion of the Genetically Altered Mutant Test Cases!

In my lectures on Software Testing I remark that there is not nearly enough research done on software testing. This summer, in my web based software engineering course we were discussing mutation testing and one of my students mentioned the potential application of Genetic Algorithms to mutation testing. After a bit of web searching I came across two articles that explore Genetic Algorithms for testing (pdf of article 1, pdf of article 2). Please share any comments you have on these articles or additional articles on this topic. Please also share any additional resources, pointers to research on software testing. Thanks.

In doing the research for this post, I came across yet another source for free educational material on the web. Samizdat Press has an eclectic collection of articles including a tutorial on LaTex and pointers to other free educational material sites. In the near future I hope to list most of these resources on my homepage.

My summer posting has been light. Hopefully you will see increasing frequency in the next few weeks. Later!

Book Learning

On Friday I gave a talk at Hobart and William Smith Colleges. It is a revision of the earlier talk I gave at Drexel and can be accessed either at my home page or here. If you are interested, one of the faculty members at Hobart and William Smith, Professor Eck, has written a great introductory text on java that is available on the internet for free! You can get it here.

Speaking of free books, I downloaded this intro calculus book by H. Jerome Keisler of the University of Wisconsin and was very impressed. A post on slashdot recommended it.

One book that is not free but is the introductory and reference text for Artificial Intelligence is Stuart Russell and Peter Norvig's, Artificial intelligence: A modern approach, Prentice Hall, second edition, ISBN 0-13-790395-2. I plan to use it in a future course and it would be a great book for anyone wanting to understand the current issues in AI.

In a later post I will survey some of the more general places on the web for free downloadable texts. I would appreciate any suggestions of worthwhile downloadable texts or comments on the texts I suggested. Later!

SCENE Conference

I gave a talk at the SCENE Conference in Philadelphia on Saturday, February 21st. SCENE stands for Student Chapter Event - Northeast, and it was an all day event where the ACM student chapters from universities in the Northeast met and attended presentations on a central theme (AI this year). This is the abstract:

Artificial Intelligence has been an active part of the computing industry for three decades. I will provide a personal view of AI in industry over that time, highlighting both the engineering and research aspects, compare it to AI at the university and offer advice for anyone making the transition from the university to industry. I also will speculate on the future of AI in industry.

You can download the pdf here. Rather than starting a new blog for AI I am adding it to this blog under the category artificial intelligence.