Is Computer Science Dead?(bcs.org)
bcs.org
Is Computer Science Dead?
http://www.bcs.org/server.php?show=ConWebDoc.9662
8 comments
We really need a down button on the links. So many of them suck.
What is left that is interesting to do in CS? Why is it "completely ridiculous" to suggest that everything interesting in CS has been done? Disprove it by example.
If I could prove what could still be done in CS I'd do it myself and become rich and famous ;)
As for something I saw recently that blew my mind was video of a presentation at google. The premise was figuring out how to have a system automatically compress, decompress, and predict based on arbitrary data. I wish I could find the link. Very neat stuff.
No matter what, there's always more to learn and discover.
As for something I saw recently that blew my mind was video of a presentation at google. The premise was figuring out how to have a system automatically compress, decompress, and predict based on arbitrary data. I wish I could find the link. Very neat stuff.
No matter what, there's always more to learn and discover.
AI, Higher level languages. Greenspun's idea to have a formal description for websites so you don't need to keep reinventing the wheel, Quantum computers, chemical computers, gesture interfaces, robotics, and so on.
Use of parallel processing is still in its infancy. Virtual reality is still in its infancy. Speech processing is still in its infancy. Wait ... was that comment a joke?
I'm wondering if this reply is a joke because none of those areas are still in their infancy, especially parallel processing and speech processing. If you really meant to say that their potential hasn't been realized, then is it possible that expectation levels were too high to begin with? Same thing with higher level programming languages. How long have people been working on ML, Haskell, LISP, etc? What you really mean is natural language programming languages, like English, Chinese, etc. as programming languages.
The flip side is that some problems that were thought to be easy to solve may, in fact, be unsolvable (practically or theoretically). Computer vision is a good example of this.
I'm not trying to start a war or play devil's advocate for the sake of playing. My comments come from a real place. I have a PhD in computer science and was a professor in the field at a top university in the US. As sobering as it is to think about, I have serious concerns about CS because the problem landscape looks a little flat. The first mountain was computation (e.g., microprocessors), then usability, then networking/communications. What's next? I believe a big breakthrough in established fields like AI, VR, etc. are low probability using existing computational models, the weak proof being that a lot of very bright people have been working on these hard problems for a very long time and advances appear to be small and incremental. I don't know if QC will hold any promises in this area, maybe so.
The flip side is that some problems that were thought to be easy to solve may, in fact, be unsolvable (practically or theoretically). Computer vision is a good example of this.
I'm not trying to start a war or play devil's advocate for the sake of playing. My comments come from a real place. I have a PhD in computer science and was a professor in the field at a top university in the US. As sobering as it is to think about, I have serious concerns about CS because the problem landscape looks a little flat. The first mountain was computation (e.g., microprocessors), then usability, then networking/communications. What's next? I believe a big breakthrough in established fields like AI, VR, etc. are low probability using existing computational models, the weak proof being that a lot of very bright people have been working on these hard problems for a very long time and advances appear to be small and incremental. I don't know if QC will hold any promises in this area, maybe so.
If speech processing is not in it's infancy, why can't my laptop tell how I'm feeling by listening to my voice? I seriously doubt you think there's no way to accomplish that.
Perhaps we are not talking about the same thing? The tone of the article was that students were being turned away from CS as a major because they felt nothing new was happening. In response to that I'm talking more about applications of these technologies. The applications of parallel processing haven't even begun to be tapped out. Same with the applications of speech processing.
I once used gp to create a soccer bot for the robocup. It took hours to train it to even kick the ball on my iMac running Mandrake. With more advanced use of parallel processing I could have trained a much more sophisticated bot -- but that was not available to me at the time.
As for the theory of computation, perhaps that field is completely tapped out. I have no idea -- you would probably know better than I.
Perhaps you're right. Maybe all the programs have been written (theoretically). Perhaps there's nothing left to be done.
Then again, computer science has only been around since the mid 20th century. Chemistry and physics have been around much much longer. It just seems that it's a bit too early to call.
Perhaps we are not talking about the same thing? The tone of the article was that students were being turned away from CS as a major because they felt nothing new was happening. In response to that I'm talking more about applications of these technologies. The applications of parallel processing haven't even begun to be tapped out. Same with the applications of speech processing.
I once used gp to create a soccer bot for the robocup. It took hours to train it to even kick the ball on my iMac running Mandrake. With more advanced use of parallel processing I could have trained a much more sophisticated bot -- but that was not available to me at the time.
As for the theory of computation, perhaps that field is completely tapped out. I have no idea -- you would probably know better than I.
Perhaps you're right. Maybe all the programs have been written (theoretically). Perhaps there's nothing left to be done.
Then again, computer science has only been around since the mid 20th century. Chemistry and physics have been around much much longer. It just seems that it's a bit too early to call.
If the premis is that cs programs are out of date, then yes, the article has a point. If the premis is that everything in software has been "done" and there's nothing interesting left to do, then it is completely ridiculous.