Over the past years, Yahoo! Research has become a powerhouse of research in algorithmic economics (as well as other areas). Sadly, their major basic research push that started with Noble intentions to win the Nobel prize (presumably in economics) has been facing serious difficulties for a while. First there were rumors about cancelled job interviews, then Prabhakar Raghavan jumped ship, and by the beginning of April it was clear that a tsunami was sweeping through Yahoo! Research. Personally I’ve had many great interactions with esteemed Yahoo colleagues, especially in the past year, and in fact I am the proud recipient of one of their space pod chairs! (Its delivery was an interesting experience, which started with a phone call from the delivery company about a mysterious 300 pound “machine”, and continued with the arrival of a huge, impregnable crate that resisted efforts to open it for two days.) So the Yahoo situation was on my mind.
This week I was delighted to read the news that 15 of the Yahoo! NYC researchers have been assimilated into the benign MSR collective, and in fact will form the basis of the new MSR NYC lab. The first stories only mentioned by name three extremely prominent researchers: Dave Pennock, Duncan Watts, and John Langford. Today Dave disclosed a few more names: David Rothschild, Sharad Goel, Dan Goldstein, Jake Hofman, and Sid Suri. Wow, what a catch for Microsoft! Interestingly, much like the recently founded MSR Herzeliya lab (which is led by Moshe Tennenholtz), the MSR NYC lab will report to Jennifer Chayes at MSR NE, with Dave Pennock in a leadership role.
I have it on good authority that, going forward, Yahoo labs may actually increase their investment in sponsored research, which I think makes sense. So the main uncertainty at this point seems to be the fate of the Santa Clara lab. At least two of the prominent SC-based members of the Microeconomics and Social Systems group, Mohammad Mahdian and Preston McAfee (the group’s leader) have recently become nooglers.
In any case, it seems that the dust is settling, and overall the glass is half full: this mayhem has led to the creation of an exciting new lab and new opportunities for algorithmic economics. Dave, Duncan, Sid, et al. — congratulations and good luck!
Basic research is high-risk and it can take over 10 years to bear fruit. The ecosystem that allowed Bell Labs to emerge and prosper no longer exists in the United States. Yahoo is a sinking ship, and killing all that is non-essential is a necessity.
Yahoo lost some outstanding people, but the changes may be the right thing for the company, as Dave Pennock pointed out (http://blog.oddhead.com/2012/05/05/goodbye-yahoo/). However, I wouldn’t necessarily jump to general conclusions about the ecosystem: How is the MSR culture different from Bell Labs? (Except for the focus on computer science.)
Bell Labs spawned the transistor, information theory, the communications satellite, the vocoder, UNIX, the C programming language, etc etc etc. Although MSR has produced some very interesting work, its impact has been a few orders of magnitude lower than Bell Labs’. Please do note, however, that I am not blaming MSR researchers; after all, it is not their fault that they were born a few decades too late and missed a golden era.
Bell labs was founded in 1925, and had a long time to spawn stuff. MSR was founded in 1991. Give them a few more years… I would say that the current era is as golden as any for technological innovation.
Good point. But, let us keep in mind that in the three decades after Shannon published his seminal 1948 paper, the field of communication systems advanced enormously. Real progress in electronic digital computers took less than 10 years. It took less than 20 years to go from the invention of the integrated circuit to the Intel 8086. The Manhattan Project yielded formidable results in less than 4 years. By contrast, it’s been almost three decades since people started talking about quantum computing, and the QC community has almost nothing “real” to show. Nuclear fusion has been a “it’s only 20 years away!!” kind of problem for decades.
I do not know why the mere suggestion that not all fields are equally fertile elicits such strong emotional responses from some academics. Some problems are hard. Some problems are impossible. Some problems require tools that have not yet been invented to be solved.