Cornell Blog: An unofficial blog about Cornell University

This robot is given nothing about its own abilities, and only the goal of getting from point to point b. It does pretty well:

The idea behind the robot is one of exploration and evolution. First, it builds up a set of possibilities of how its parts are arranged. Then, it builds up sets of possible commands to send to those parts. The commands are selected explicitly to try to maximize the differentiation of its internal models of itself. Once the robot has figured out what its made of and how those parts work together, it can attempt to move or accomplish a task. While this “learning stage” is limited to only 16 iterations to prevent a sort of machine epilepsy as it tries out badly performing permutations, it could make a useful fall-back algorithm for accomplishing mechanical tasks in a damaged scenario.

The Cornell News has more information with their article Cornell robot discovers itself and adapts to injury when it loses one of its limbs. What a title.

What happens when you take the “science” or “engineering” out of Computer Science and add in a few warm fuzzies? You get MIT’s new joint program in Web Science with the University of Southampton, Britain.

According to the New York Times, the program is all about what people do with computers:

“The Web isn’t about what you can do with computers,” Mr. Berners-Lee said. “It’s people and, yes, they are connected by computers. But computer science, as the study of what happens in a computer, doesn’t tell you about what happens on the Web.”

As Threadwatch points out, no one is going to respect this useless major. For example, one area of study might be a strange idea of privacy:

Privacy, for example, will be one area of research in Web science. The traditional approach to protecting privacy has been to restrict access to databases containing personal information. But so much personal information is already available on the Web, often given voluntarily on sites like MySpace and Facebook, that the old approach will not work, said Daniel J. Weitzner, technology and society director at the Web consortium.

Studying social phenomena is not science, so the major is ill-named. Also, studying how popular web sites work is not academia–it’s web surfing. If you want to understand how web or distributed systems work, you need a background in computer science and experience building them yourself, not a crappy degree from MIT. When recruiters see Web Science they should think the same thing as Did no work and partied through college.

According to the worst piece of research I’ve ever seen, watching more TV causes Autism. Nodir Adilov, Sean Nicholson, and Michael Waldman should be publicly scolded for letting this be published. Time, in a scathing article titled A Bizarre Study Suggests That Watching TV Causes Autism, summarizes:

Lo and behold, Waldman and colleagues found that reported autism cases within certain counties in California and Pennsylvania rose at rates that closely tracked cable subscriptions, rising fastest in counties with fastest-growing cable. The same was true of autism and rainfall patterns in California, Pennsylvania and Washington State. Their oddly definitive conclusions: “Approximately 17% of the growth in autism in California and Pennsylvania during the 1970s and 1980s was due to the growth of cable television,” and “just under 40% of autism diagnoses in the three states studied is the result of television watching due to precipitation.”

Ignoring the absurd problems that arise when attempting to confuse causality with correlation, the very idea that watching television, a process which occurs late in a child’s life, causes autism, a condition which is usually there by birth, is ludicrous!

This is a bad PR move on Cornell’s part. Aren’t Ivies supposed to be smart?