The opening of Microsoft Research's latest facility was celebrated today with a free one-day symposium here in Cambridge, Massachusetts. I think the symposium succeeded in its goals of showing that the research facility is an independent entity that plays by the rules of open scientific debate and funds basic research of value to society.
The facility opened in July, but the conference was scheduled today to give staff and interested attendees a chance to return from summer holidays. The location, appropriately, was MIT's Stata Center, which has a wing donated by the Gates family.
Some 200 people showed up to see and be seen, but we caught a number of interesting technical presentations from current and upcoming staff. A few highlights:
- Leading CS researcher Rick Rashid, who founded Microsoft Research 17 years ago, claimed not only that most Microsoft products incorporate his lab's research but that the lab has succeeded in prompting Microsoft product developers to think more rigorously and theoretically in their everyday work. He stressed, however, that Microsoft Research's first goal is to "expand the state of the art" and benefit all computer users.
- Ethnologist danah boyd explained some of the ways social networks oversimplified human interactions and produced stress, as well as a variety of clever adaptations, among the young people who feel a need to network there. Her descriptions of why they spend so much time were also quite sad: many are constrained from going out and leading a normal life in their neighborhoods by paranoid parents afraid of abductions, and if the teens do go out there are often no places left that are friendly to them.
- Economist Susan Athey elucidated at breakneck speed some of the unintended consequences created by various ways to measure and auction search engine advertising.
- Veteran designer Bill Buxton warned that product requirements should include the user experience--such as "Every user can learn the first 80% of the functionality in ten minutes"--and that lacking such requirements, products will continue to be plagued by bloat and interface difficulties. He said the first rule of design is to present end users with multiple prototypes, because that frees the viewers to stretch their imaginations and critique each prototype. In fact, you should ask a user to draw out his own ideal prototype.
- Algorithm researcher Erik Demaine waxed enthusiastic about the reach of computational thinking into all areas of human life, including games, agriculture, medicine, and art.
I myself apply numerous algorithms to writing and editing, one of which tells me I have already spent too much time describing individual talks. The overall message is that Microsoft's 850 researchers are having fun and making contributions to society. The Cambridge facility employs a wide range of scientists, including social scientists and biologists. Biology, a well-known Boston-area strength, is an area where more research can produce obvious social benefits; in addition (as managing director Jennifer Chayes pointed out), the computational and network elements of genetics show intriguing parallels with the computing field.
The number of publications and presentations released by Microsoft Research shows that it's an active, and even premier, research facility. But it's also a generator of many patents. Since the Baye-Dole Act, of course, so are universities. Patent-free commons don't exist any more. Microsoft could demonstrate its commitment to openness by putting patents into open source patent pools such as the Open Invention Network, which was invented partly as a defense against Microsoft itself.
But I'm happy to see Microsoft doing research, especially here in Massachusetts. The mighty elephant, grazing the savannahs of computer technology and trying to leave its footprints everywhere, has proven adept at creating an ivory tower. Congratulations to vice president Rashid, director Chayes, and everyone who put it in place.