4 ms·

I see something very similar happening in many Computer Science papers and the areas in which some CS academics have chosen to focus their time. It's not as ba

by tytso 15y ago

I see something very similar happening in many Computer Science papers and the areas in which some CS academics have chosen to focus their time. It's not as bad as what goes on in Law School, certainly, but there is definitely a divide between what is important to industry (which is often much more about engineering) and academics (who need to publish something which will get them tenure, which means it has to pass muster with senior academics in the field).

_delirium 15y ago
A difficulty is actually predicting that, though, especially if you're looking for medium-term influence rather than usable-immediately. For example, many "practically oriented" software engineers would've derided Okasaki's thesis, Purely Functional Data Structures, as pointless mathematical masturbation, until a decade later it became one of the core parts of Clojure, which suddenly (some) software engineers found useful.
tikhonj 15y ago
I think that, for CS, this is actually a good thing: "industry" in general tends to be more interested in short-term gain and things which are immediately practical--they are creating the technology of tomorrow. Academics, on the other hand, are not bound by this; they can look into creating the technology of the far future. Another important difference is that CS research also goes on in industry, resulting in more diversity. There is no fundamental problem underlying the whole field--for every researcher working on things you find useless for industry, there are probably several working on things which are useful. Finally, this may be due to my university being more pragmatic than some, but most of the research going on that I've heard about is actually immediately practical. The single largest issue a ton of professors seem to be working on is parallelism and distributed computing, which is indubitably extremely important to industry right now. Of course, I could also only be seeing this because most of the professors that teach my classes or that I've talked to are affiliated with similar labs.

by tytso 15y ago

_delirium 15y ago

A difficulty is actually predicting that, though, especially if you're looking for medium-term influence rather than usable-immediately. For example, many "practically oriented" software engineers would've derided Okasaki's thesis, Purely Functional Data Structures, as pointless mathematical masturbation, until a decade later it became one of the core parts of Clojure, which suddenly (some) software engineers found useful.

tikhonj 15y ago

I think that, for CS, this is actually a good thing: "industry" in general tends to be more interested in short-term gain and things which are immediately practical--they are creating the technology of tomorrow. Academics, on the other hand, are not bound by this; they can look into creating the technology of the far future. Another important difference is that CS research also goes on in industry, resulting in more diversity. There is no fundamental problem underlying the whole field--for every researcher working on things you find useless for industry, there are probably several working on things which are useful. Finally, this may be due to my university being more pragmatic than some, but most of the research going on that I've heard about is actually immediately practical. The single largest issue a ton of professors seem to be working on is parallelism and distributed computing, which is indubitably extremely important to industry right now. Of course, I could also only be seeing this because most of the professors that teach my classes or that I've talked to are affiliated with similar labs.