I spend a lot of time talking to people in academia, particularly in the computer science research community. I'm sorry to say, there is an emerging crisis in the field in the U.S., and that crisis is funding.
Let me preface this by saying that Microsoft Research does not take any federal funding for our research activities; we are entirely self-funded. So this is not an argument that is going to line the pockets of my employer or myself, though I certainly have a stake in the overall health of the computer science research community.
There are three major sources of government funding in the U.S.: NSF, DARPA and the DHS. NSF has traditionally been the most stable of the three; DARPA is actually a military organization and tends to follow the ups and downs of military spending, and DHS is still very new.
Within NFS, the program that funds most CS research is called CISE. Here is a link to an interesting report in the CRA's newsletter reflecting on what has happened (or not happened) to CISE over the last decade. A few highlights to catch: the rejection rate for funding requests is now reaching 85% -- and for junior faculty, it's much worse. To make matters worse, Congress actually cut the NSF budget in the FY05 budget passed a couple of weeks ago.
What's going on here? Well, several things. First of all, there was a significant spike in the number of CS faculty in the late '90s into the early part of this decade -- thus increasing the number of proposals. You could argue that it is in fact an unsustainable number in the post-dotcom era, and it is in fact another bubble that will burst at some point.
But let's step back a moment and take a look at the demand, from a point of view of our economy and our national priorities. From the economy's point of view, the big drivers of growth are computing and life sciences; and the life sciences are being driven by advances in computing. So in many ways computing is the goose that lays golden eggs.
In terms of national priorities, you can read a lot into the FY05 budget. (check out figure 5) But what's more sobering is looking into the details of the homeland security efforts across NSF, DARPA and DHS. Here's a link to a PITAC presentation on just that.
In the FY04 NSF Cyber Trust program, there was only $31M of research grants, which funded about 8% of proposals (about 25% were deemed worthy of funding by review panels). In the DHS Science and Technology budget ($1.069B) a mere $17.8M is earmarked for cyber security. DARPA is an even worse mess: they have classified their cyber security research program, which will prevent most academic researchers from participating (including most of the foremost cybersecurity experts) and will also prevent tech transfer into the commercial sector. Further, we have heard rumors that DARPA is about to change their rules and discontinue altogether research contracts directly with universities, requiring a commercial entity as a go-between.
There is broad agreement that cyber security is one of the most important issues facing us today, and the risk of cyberterrorism is quite high. The PITAC report I mentioned earlier gives some eye-opening statistics on how bad the problem is in the private sector too. This is a problem that we cannot ignore... and yet that is exactly what our government is doing, in favor of old-school military and physical security. They just don't get it.
This crisis is about CS research, but not just about that. It's about cyber security, but not just about that either. It's also about jobs. If the CS community implodes, it takes with it our ability to fill the significant demand for highly-trained technical employees in our country. India and China are investing very heavily in their science and technology research community and academic infrastructure; and we, in turn, are de-committing from ours. Despite what you might read in the papers, highly skilled technical jobs are generally not being offshored to replace existing positions here; however, there are companies starting to look offshore because they want to grow their business and they simply can't find enough qualified people here. And we are on a trajectory now where this problem is going to get far, far worse. In 10 years, India and China will be science and technology powerhouses... and the U.S. will be squandering its long-held advantage.
What to do? First, fund basic research in science and technology. Countless studies have shown that strong economic centers always grow up around top-tier research universities (Silicon Valley, Seattle, Boston, Virginia, San Diego, Cambridge, Beijing, Shanghai, Tokyo... the list goes on). Second, get a critical mass of effort and funding on cyber security, before it's too late. Third, relax the student visa requirements so that the best and brightest from other countries can come here to study; it actually strengthens our universities, breeds fresh ideas, and eventually draws many of them into our domestic industries which makes us more competitive -- and there is such a high demand for skilled science and technology workes here that it really doesn't take away jobs from American citizens.
This is a huge crisis that is looming over us. We need to let our elected representatives in D.C. know that the time to act is now, before we kill the goose that lays the golden eggs.
10:22:11 PM  ; ;
|
