5 ms·

Is Computer Science Dying?

27 points by danw 19y ago

ivankirigin 19y ago
Now that computer scientists have discovered that the statistics and probability research of the past few decades completely applies to any interesting problems, there should be at least a few more good years left. Regardless, computer science would never actually die. It would just deadlock or attempt to solve the halting problem.
amichail 19y ago
It's amazing that computer scientists never consider the possibility that they are wrong -- it's always students who are misguided. In my view, computer scientists are wrong. They have limited the scope of the field to exclude application-level creativity. And in so doing, they have severely restricted the potential of the greatest invention of all time.
- programnature 19y ago
  amen to that. one might say they've prematurely optimized on what is interesting about computation, before we had enough practical experience to really know.
- greendestiny 19y ago
  There's no way a field of study should be telling you what to do with your study. Computer science has always been about application-level creativity, its about studying the methods by which computers can be made to do useful things. Coming up with these useful things is your job.
  - amichail 19y ago
    So how come we have people studying art, music, and writing in university? And note that such studies do not normally involve learning about the chemicals used in the paint, the electronics used in sound equipment, or the software architecture of the word processor. At the very least, students should be taught how to be social scientists so that they could figure what sorts of apps are likely to succeed in a particular environment.
    - brent 19y ago
      I believe what you're referring to then is the need for an increase in informatics programs (like at my school www.informatics.indiana.edu). However, I should stipulate that this is use of a computer to do tasks related to information, not computer science per se.
      - amichail 19y ago
        You don't need to call it computer science. However, as far as I know, such undergraduate degrees don't exist.
        brent 19y ago
        You didn't look at Indiana. http://www.informatics.indiana.edu/academics/sample_curriculum.asp http://www.informatics.indiana.edu/academics/sample_curricul... This is the undergraduate informatics curriculum which, as you'll notice, strictly includes social science courses throughout the first two years.
    - greendestiny 19y ago
      I'm utterly unconvinced that formal studies in art and writing produce great artists. While I imagine it gives a lot of interested people, some of whom with great talent, an opportunity to explore their skills, the research and accumulated bodies of knowledge are mostly a cataloging of the works of talented people and not a how-to guide. Formal studies in computer science teach the accumulated knowledge of building systems with algorithms. Its like learning the language, the grammar and vocabulary, a great writer needs. You're right, it is like learning about the chemicals in the paint, or more importantly the properties of your paint, your canvases and the brushes and techniques other artists have used. My point is that that is the useful knowledge. Don't confuse universities offering a socially acceptable outlet for aspiring artists as a model for study and learning. You can probably build a great application without being a great computer scientist, but why would computer science want to be one of these sham fields of study when their are real tools and knowledge to be discovered.
      - 19y ago
- comatose_kid 19y ago
  Nothing wrong with understanding computer architecture/algorithms/os/etc before you exercise your application-level creativity. Anyone learning a sport has to master basic things before they go on to the advanced (eg in volleyball, you learn how to bump/set/spike before you learn how to run offensive plays). I think the same applies to CS.
  - umjames 19y ago
    But at least you do get to learn how to run offensive plays while still learning volleyball. Personally speaking, in my 5 years of college (Drexel University from 1996-2001), most classes were 80-100% theory and 0-20% programming. None of it was application-level programming until the Software Engineering track in senior year. And you didn't have to have a full working program to pass either; it was enough to just show you made an effort. I'd rather see a university that promoted the idea of letting self-motivated students do their own thing instead of just turning education into a checklist of required classes and credits.
sabat 19y ago
Only like the phoenix.
wozer 19y ago
> One solution proposed for this lack of vision is that of renaming the subject to "informatics." Renaming doesn't create visions but it should be noted that this is the usual term in French, German, Italian, Norwegian etc.
nradov 19y ago
The biggest problem in the field is still that most universities don't offer separate Computer Science and Software Engineering majors. While there is substantial overlap, those are two distinct fields. It's much like the relationship between Physics and Mechanical Engineering. The IEEE is doing some good work to establish a standard curriculum for Software Engineering and get it implemented at more schools. That would provide a more attractive option to students who want to create software but don't enjoy the theoretical and mathematical side as much.
michaelneale 19y ago
I hope not. The mathematical aspects are about to get a whole lot more important to us.

27 points by danw 19y ago

Now that computer scientists have discovered that the statistics and probability research of the past few decades completely applies to any interesting problems, there should be at least a few more good years left. Regardless, computer science would never actually die. It would just deadlock or attempt to solve the halting problem.

amichail 19y ago

It's amazing that computer scientists never consider the possibility that they are wrong -- it's always students who are misguided. In my view, computer scientists are wrong. They have limited the scope of the field to exclude application-level creativity. And in so doing, they have severely restricted the potential of the greatest invention of all time.

programnature 19y ago
amen to that. one might say they've prematurely optimized on what is interesting about computation, before we had enough practical experience to really know.
greendestiny 19y ago
There's no way a field of study should be telling you what to do with your study. Computer science has always been about application-level creativity, its about studying the methods by which computers can be made to do useful things. Coming up with these useful things is your job.
- amichail 19y ago
  So how come we have people studying art, music, and writing in university? And note that such studies do not normally involve learning about the chemicals used in the paint, the electronics used in sound equipment, or the software architecture of the word processor. At the very least, students should be taught how to be social scientists so that they could figure what sorts of apps are likely to succeed in a particular environment.
  - brent 19y ago
    I believe what you're referring to then is the need for an increase in informatics programs (like at my school www.informatics.indiana.edu). However, I should stipulate that this is use of a computer to do tasks related to information, not computer science per se.
    - amichail 19y ago
      You don't need to call it computer science. However, as far as I know, such undergraduate degrees don't exist.
      - brent 19y ago
        You didn't look at Indiana. http://www.informatics.indiana.edu/academics/sample_curriculum.asp http://www.informatics.indiana.edu/academics/sample_curricul... This is the undergraduate informatics curriculum which, as you'll notice, strictly includes social science courses throughout the first two years.
  - greendestiny 19y ago
    I'm utterly unconvinced that formal studies in art and writing produce great artists. While I imagine it gives a lot of interested people, some of whom with great talent, an opportunity to explore their skills, the research and accumulated bodies of knowledge are mostly a cataloging of the works of talented people and not a how-to guide. Formal studies in computer science teach the accumulated knowledge of building systems with algorithms. Its like learning the language, the grammar and vocabulary, a great writer needs. You're right, it is like learning about the chemicals in the paint, or more importantly the properties of your paint, your canvases and the brushes and techniques other artists have used. My point is that that is the useful knowledge. Don't confuse universities offering a socially acceptable outlet for aspiring artists as a model for study and learning. You can probably build a great application without being a great computer scientist, but why would computer science want to be one of these sham fields of study when their are real tools and knowledge to be discovered.
    - 19y ago
comatose_kid 19y ago
Nothing wrong with understanding computer architecture/algorithms/os/etc before you exercise your application-level creativity. Anyone learning a sport has to master basic things before they go on to the advanced (eg in volleyball, you learn how to bump/set/spike before you learn how to run offensive plays). I think the same applies to CS.
- umjames 19y ago
  But at least you do get to learn how to run offensive plays while still learning volleyball. Personally speaking, in my 5 years of college (Drexel University from 1996-2001), most classes were 80-100% theory and 0-20% programming. None of it was application-level programming until the Software Engineering track in senior year. And you didn't have to have a full working program to pass either; it was enough to just show you made an effort. I'd rather see a university that promoted the idea of letting self-motivated students do their own thing instead of just turning education into a checklist of required classes and credits.

sabat 19y ago

Only like the phoenix.

wozer 19y ago

> One solution proposed for this lack of vision is that of renaming the subject to "informatics." Renaming doesn't create visions but it should be noted that this is the usual term in French, German, Italian, Norwegian etc.

nradov 19y ago

The biggest problem in the field is still that most universities don't offer separate Computer Science and Software Engineering majors. While there is substantial overlap, those are two distinct fields. It's much like the relationship between Physics and Mechanical Engineering. The IEEE is doing some good work to establish a standard curriculum for Software Engineering and get it implemented at more schools. That would provide a more attractive option to students who want to create software but don't enjoy the theoretical and mathematical side as much.

michaelneale 19y ago

I hope not. The mathematical aspects are about to get a whole lot more important to us.