Manufacturing Innovation
There's a fascinating New Yorker article on Nathan Myhrvold's current company, Intellectual Ventures. One of the company's core activities involves getting lots of smart people together in a room for a few days, brainstorming, and patenting the resulting ideas. They're kind of a patent troll outfit, but they do seem to be generating some genuinely interesting ideas and pushing at least some of them into commercialization.
Reading the piece brought to mind the various efforts by NSF and others to push interdisciplinary research. IV's efforts are explicitly interdisciplinary, but they have an extra ingredient that I think is absolutely crucial: they go and talk to people outside of academia who are trying to solve real problems. NSF has had a handful of initiatives that involve getting academics out of the ivory tower - industrial postdocs in math come to mind - but not very many. The real world is a source of a lot of great inspiration for research, and not just applied research. For example, some of my own work on wireless communications led me to some deep problems in number theory (distributions of digits in binary representations of elements of the Cantor set; I never got very far, so it's wide open). I think funding agencies could get a lot of scientific (not to mention political and economic) value from encouraging greater interaction between academics and people in industry.
The researchers involved in IV have rather interesting backgrounds, which the article's author credits in part for their ability to generate novel ideas. Their resumes aren't of the form, "got PhD from X, did postdoc at Y, became faculty at Z, wrote N papers;" rather, these are people who have walked across Texas on a whim / have leveraged careers in astrophysics into paleontology, etc. Two (Myhrvold and Jung) were former bosses of mine at Microsoft, so let's hope I'm on a good track.
-
on Tue, May 13, 12:05PM
Geoff - Thanks for this very interesting article. It is interesting how the whole academic gestalt and economy is oriented toward the "individual discoverer". Rarely, I have seen PIs who assemble a great team of partners around them and the group ends up being enormously productive. But, even in those environments, there is a single PI who is "boss" (and, often first author on all the papers!).
The article also discusses the surprising frequency of "simultaneous invention" which suggests that good ideas are as much the product of their time as they are the product of the mind of a "genius."
A variant on the I.V. model is the emerging concept of "crowd-sourcing" - using a vast, disseminated groups of people to edit reference materials (Wikipedia), sieve through content for the few jewels (Digg), and numerous sites and organizations that use rating systems (Zagat, EBay, etc...).
Science Mag (John Travis, author, Science 28 March 2008: 1750-1752) had a recent article about a company called InnoCentive that is using crowd-sourcing for invention. Companies post technical challenges and a bounty and anyone can submit a proposal for solving the problem. They have quite a few successes under their belt. It again illustrates the idea that great ideas are out there - but connecting to the people who have them is the challenge. Again, this also belies the academic concept that there are a few much-vaunted intellectual "leaders" (called professors) who we rely upon to invent everything - and a sea of minions who set about implementing the vision of their genius bosses.
My experience with two start-ups has shown me that "invention" is typically not the barrier to financial success in a company. In many ways, invention is the most enjoyable part of the process. But turning an invention into a product, then a set of products, then a growing stream of revenue, requires almost the opposite set of skills: single-minded focus, banal (almost mechanistic) discipline in execution, and patience. Long after the inventor has grown bored and wandered off to collect more "daisies," the "do-ers" are left behind to make something real out of the lofty ideas.
Another data point: Lowell Wood, a former physicist at LLNL, was a main feature of this article. Lowell is no doubt a brilliant and creative individual. But if one cooly examines the actual "products" that came out of LLNL when Lowell and his O-division group were active they were, for the most part, lofty, impractical, and unrealized. The largest product of course was the X-ray laser -a technical fiasco of monumental proportions that consumed vast millions of federal dollars (see William Broad's book "Teller's War" for a detailed account of this).
And all this reminds me of a fortune cookie I got a few years back: "Those that have imagination without learning have wings but no feet."
-
on Wed, May 14, 10:05AM
Hi Peter--
Yeah, I agree that the problem is rarely a lack of ideas. Check out this great essay by Paul Graham, the organizer of Startup School: http://www.paulgraham.com/ideas.html It's incredibly challenging (and interesting / lucrative / frustrating / risky) bringing ideas to market.
I think that it's in the getting-to-market phase of innovation where the US has a lot of advantages over potential competitors, which is why, despite the lackluster educational performance that many bemoan, we continue to do well economically. But that doesn't mean that the process can't be improved, and I think funding agencies could have a useful impact - hybrid science / business programs, structured industrial internships, and so on.
As for your comments on Wood, my impression of Myrhvold (outgoing head of Microsoft Research when I was there) and of Jung (my boss's boss my first year at MS) is pretty similar. They're smart guys, and they're creative, but they have spent a bit too much time thinking and not enough doing. MSR hasn't really produced a huge amount of value for MS given how much money they have put in, and Myrhvold was responsible for some of the more expensive and useless components of MSR. And Jung's project that I worked on was a very expensive failure - a really ambitious but wildly impractical competitor to WinFS (which also managed to crash and burn a few years later). So I am guessing that IV will remain primarily a patent troll firm with great PR.
Ok, back to doing...
