Does Society Need Computer Science Research? Shrisha Rao University of Iowa srao@es.uiowa.edu Over several years of study and research in computer science, I have found myself asking this rather embarrassing question-of myself and, increasingly often, of other computer scientists, whose reactions range from ponderous thought at its profundity, to openlyexpressed amazement at its silliness. Certainly, a flip affirmative answer can be given, justified by the reasoning that technical advance is the root cause of improvements in society, that computers are the mainstayof technical advance in the present day, and that computer science is verily the branch of learning that makes computers happen. However, as with most flip answers, this one is quite unsatisfactory upon deeper investigation. For one thing, computer science is mostly about research, and yet, unlike in medicine for example, research breakthroughs in computer science rarely mean anything to society in their immediate future. It needs to be kept in mind that computer science is not I T - - and increasingly, IT is where the action is, and it is even felt by some that computer science is to IT what astrologyis to astronomy. The general public does not really understand-or care-what computer scientists do, and if asked, its answers are most likely to betray a complete lack of understanding of the difference between computer science and IT, At the same time, computer scientists have hurt their cause by claiming too much, by not seeing their good ideas to fruition, and by producing too few artifacts that caused any splash. The bold and brassy claims about AI made in former decades are now reduced to dim memories and the stuffofthe occasional Hollywood offeringonly minute achievements that often amount to little are to be seen. Good research such as the body of work on functional programming languages goes un-utilized while ad hoc and misconceived stopgaps such as languages like Perl and Python find widespread use in practice. The World Wide Web, certainly the biggest contemporary cause of attraction to computers for society at large, was created by physicists working on something else and subsequently co-opted by industry- computer scientists contributed nothing of note. Some very unpleasant consequences thus follow ~br computer scientists. Libraries and bookstores are getting less and less likely to carry monographs on serious computer science, preferring instead to stack their shelves with a wide array of buzzword-for-dummies books. Academic departments are having trouble getting, and holding on to, enough doctoral students to carry out serious research; many graduate students prefer to leave before the Ph.D. to get a wellpaying but essentially mindless programming job in the IT sector. The Communications of the ACM, which for long decades was a repository of some of the finest research papers in the field, is now little more than an opuscular IT trade mag. The threat to traditional computer science from IT and its promise of instant gratification is therefore not to be underestimated. The recent downturn in the economy may yet prove to be a blessing for computer science, as it may, for instance, help convince some prospective computer scientist that it is not a bad idea after all to spend five or six years of his life getting a doctorate, rather than picking the alluring alternative of forming or joining a start-up with the promise of making a pile instantly. All this brings us back to my question-what exactly is being achieved, or likely to be achieved, by the furtherance of computer science research? For many computer scientists, the practice of their subject is like breathing-something to be done without thinking and needlug no justification. There is also a mystical faith that whatever they do is somehow pertinent and useful, and will lead to an improvement in the human condition, sometime, somewhere. In fact, however, computer science is too insular, largely locked into the paradigms of past decades, completely inattentive to changed circumstances and refusing to effectively participate in and control the fluid and evolving realities of the present day. Take for example the widespread and nonsensical scare caused by the so-called Year-2000 bug-not a single computer scientist of any repute believed it, yet the hype reached monstrous proportions and caused society billions of dollars that could have been better spent, all because computer scientists did not care to step forward to correct, or at least to contest, the false impressions created by people with too little understanding. No discipline survives for long without public understanding and support of its objectives. And while the confusion of computer science with IT may be thought of as a short-term boon inasmuch as it tends to falsely credit computer science research with the results achieved by industry through incremental engineering advances, it is too much to expect that such a state of affairs will persist for very long. There is a dire need to clearly define the role of computer science in the changed circumstances that now obtain, forsaking or modifying the assumptions that have been made standard in past decades. Although it would be too bold to say at this time that classical computer science research is a dying art form, it is surely necessary for computer scientists to take note of their circmnstances and make an effort to determine their role in society, rather than let it be determined for them by chance or by the flow of events that they make no effort to participate in. Computers and Society, September 2OO1

/lp/association-for-computing-machinery/does-society-need-computer-science-research-Nogf2sw6we