I have a confession: The last few days I've been experiencing some unusual symptoms: My muscles are loose. My speech has slowed. I'm sleeping deeply and without interruption. My skin has acquired an odd, ruddy cast generally associated with people who don't spend their days under florescent lights. Oddly enough, the last time I felt like this I'd also just returned from a two week vacation. It's possible I'm in a state of blissful relaxation, though I'm too unfamiliar with that feeling to make a positive diagnosis. Besides, like all afflictions (good and bad), it's sure to pass quickly. In fact, I can detect that familiar quickening of the pulse just pondering everything I have to do before the Crowdsourcing book comes out next month.
But enough with the diaristic blather. Kudos to David Cohn for, well, filling my shoes wouldn't be entirely accurate. Let's say instead I hope to fill his. In the meantime, I encountered a few choice bits since returning from my Catskills retreat.
First, the New York Times Science section published an excellent piece on InnoCentive and other efforts to apply crowdsourcing to scientific research. Most of the information on InnoCentive will be old hat to readers of this blog, but it also showed the extent to which the strategy is going mainstream:
The approach is catching on. Today, would-be innovators can sign up
online to compete for prizes for feats as diverse as landing on the
Moon (space.xprize.org/lunar-lander-challenge) and inventing artificial meat (www.peta.org/feat_in_vitro_contest.asp).
This year, researchers at the Howard Hughes Medical Institute and the University of Washington
began recruiting computer gamers to an online competition, named
Foldit, aimed at unraveling one of the knottiest problems of biology —
how proteins fold (fold.it).
I was especially intrigued by the Foldit project. I've always been careful to highlight the connection between such distributed computing efforts as SETI@home and crowdsourcing, but Foldit has a direct connection. It evolved out of Rosetta@home, which used the spare cycles on people's PCs to perform the processor-intensive calculations necessary to determine the shapes of proteins. "In Foldit, players will compete online to design proteins, and researchers will test designs to see if they are any good." The idea isn't to recruit scientists, it's explicitly aims to recruit non-scientists who might have novel ideas of how to construct a protein. As Zoran Popovic, one of the computer scientists at UW running the project notes, "Our ultimate goal is to have ordinary people play the game and eventually be candidates for winning the Nobel Prize."
In the coming months I'll be using this blog to update and correct my book (which is, alas, out of my reach as it winds its way through the printing process). In the book I note that Congress had debated whether the government should be offering prizes, a la InnoCentive, for creating solutions to thorny problems for the public good, such as creating a cheap Malaria vaccine and creating viable forms of alternative energy. What I hadn't come across is the fact that last year no less an authority than the National Research Council recommended that the National Science Foundation—the agency that provides most government funding to scientific research—offer prizes of $200,000 to $2 million in "'diverse areas' as a way of encouraging 'more complex innovations' addressing economic, social and other challenges." This signifies a major vote of confidence for the use of crowdsourcing in science. The full report can be found here (for $18.90). I wish I'd had it when I was writing the book.
One other passage caught my eye. Dwayne Spradlin, president and CEO of InnoCentive, notes that "more than a third" of InnoCentive solvers have doctorates. My immediate thought on reading this was: only a third?!? I'd always assumed that InnoCentive was one example of crowdsourcing that was resistant to the influx of amateurs that characterizes the phenomenon in general. I thought it an anomaly in this regard, but given that almost all professional science is being conducted by people with PhDs or those working on attaining one, this factoid indicates that maybe InnoCentive is as much an amateur network as, say, iStockPhoto. If anyone at InnoCentive's reading this, could you offer us more insight into the demographics of your solvers? How many of the non-PhDs are coming up with successful solutions?
On the more trivial side, I couldn't help but notice that the article seemed to avoid applying the term crowdsourcing to InnoCentive or the other initiatives in the piece. The reporter uses the term open source science instead. That works for me, though given the extent to which the term has entered the lexicon, it seems like an omission. Especially since the paper included it in a round up of new words of 2007. But what was striking is that in the same issue the Times did use crowdsourcing in a headline to a wire piece about a Finnish community, called "eCars—Now!" that is trying to apply a Wiki approach "to start converting used petrol-fuelled
cars to electric ones."
Believe it or not, this post was supposed to be a quick collection of links. And I couldn't get past the first link without writing a short essay. David, how do you do it? Find out tomorrow when I attempt a more convincing imitation of my erstwhile guest blogger.
