This article is published in the November 2011 issue.

A Shifting Perspective on Computing’s Future?

As all of us are acutely aware, recent turmoil in the financial markets, near gridlock on Capitol Hill, uncertainty about long-term fiscal planning in Congress, and concerns about hesitancy in the job market are all converging to create a somewhat daunting perspective for computing research, and especially for funding for computing research, in the next few years.

While the CRA Government Affairs group has been exceptionally active in promoting the critical role of computing research to Congress—offering testimony from computing community leaders to a range of Congressional committees, including a captivating argument to Congress showing how a now commonplace, disruptive technology like the iPad was deeply dependent on a wide range of federally funded research that preceded it—it is uncertain whether this message will overcome other forces and concerns in Congress. That implies considerable uncertainty about likely levels of federal funding for computing research.

This potential downturn in federal funding comes at a time when many institutions are seeing large-scale increases in student interest—especially at the undergraduate level, though these surges will reach the graduate population in a few years. A very informal sampling of some major computer science departments shows jumps in undergraduate majors of 20 percent to 30 percent last year, dramatically building on steady increases in enrolments over the past two to three years. So we may be headed towards a worrisome collision—an increase in student population and student interest, the emergence of a range of exciting new areas of opportunity, confronting a significant reduction in research funding, and a potentially uncertain prospect for job opportunities.

So what options are available to the computing community? One could, of course, argue that we should not be so dependent on federal funding, but that simply leaves open the question of where to find other sources. One possibility is closer ties with industry, and indeed many of our major industrial partners have stepped up both their internal support for research and their outreach to academic centers. But companies naturally take a shorter-term, more applied view of research, and while these partnerships are very valuable, it is rare for industrial funding to fully substitute for federal funding of foundational work. The growth of interdisciplinary research within the computing community not only broadens the impact of computational methods, it also widens the set of funding sources, as witnessed by the growing number of computing researchers with NIH funding.

But even with actions being undertaken to encourage broader funding support for research, the community may have to face a fundamental challenge—that we will have less support for graduate students at a time when the output of the pipeline is increasing.  And if that is correct, it suggests that the field may need to rethink some of its priorities.  How would research centers cope with a reduction of from 10 percent to 20 percent in the number of graduate students? Does this dramatically change the kinds of research projects we undertake?  In particular, would this have the potentially troubling side effect of pushing the community towards less risky, more near-term research? Are postdoctoral positions really a useful component of the research pipeline if there is an unclear trajectory towards academic research positions at the end of such positions?  And finally, should we be rethinking the advice we give to undergraduates who are considering their options?

I certainly don’t want to suggest that we discourage great students from pursuing graduate school and gaining the deeper training needed to pursue research careers.  But perhaps we have an obligation to think about alternative, yet equally challenging, career trajectories, and to articulate these choices to students. This would include other fields where computational thinking, algorithmic discipline, and analytic problem-solving are key elements to success—and would include finance, health care, large-scale systems management and delivery, and many others. And, in my view, it would include more explicit support for entrepreneurial trajectories.

As we have seen in the past, disruptive changes in jobs creation, in technological innovation, and in definition of entire new fields of inquiry (such as social media) can come from entrepreneurial forces. And concomitant with that perspective is the observation that some of our students may be much better suited for such endeavors than traditional graduate school, through a combination of personal temperament, ingenuity, and modes of thinking. As a consequence, it may be important for institutions to consider whether they are fully empowering such student opportunities. While there is no specific formula for supporting entrepreneurship, many departments are experimenting with a range of tools: subjects within the discipline on entrepreneurial successes and failures and exposure to recent examples; business plan competitions that encourage team building between engineers and management practitioners; innovation sandboxes that provide opportunities for students to explore ideas and build networks of connections; technical competitions; and hackathons.

Each department must address its own particular needs and its own cultural constraints; however, perhaps this is a good time for a department to think about how, within those constraints, it can seek ways to broaden the scope of opportunities for its students, and strengthen its base of support in advance of potentially tumultuous times.

Eric Grimson is Chancellor of MIT, the Bernard Gordon Professor of Medical Engineering, and Professor of Electrical Engineering and Computer Science at MIT.

A Shifting Perspective on Computing’s Future?