Spent the day at Science 2.0 down at the Mars Centre. Kinda freaky how many people I am getting to know at these things. Greg is, as was mentioned, definitely a node in the connectivity graph. The *P marketing department did a presentation, I've never personally looked into the system, but the few quick snippets on-screen looked elegant in an "I don't see how that extends" way. May be some other abstraction I didn't notice.
Lady from Stanford/MIT/Yale had an interesting survey on why scientists weren't going for Open Research... my impression was that mostly it came down to "too much work", with a few straw-man fears thrown in. Interesting slide pair showing the "blue versus red" breakdown of motivations for open versus closed reasoning (spoiler, altruism and idealism tends to motivate sharing).
Presentations from the polymath group and an open lab were interesting practical applications of the idea of doing scientific investigations out in the open. Titus was more of a "we should be doing more testing" talk. One thing that got mentioned (and I've forgotten by whom) was the question of how to motivate academics to contribute to/curate Wikipedia (or similar repositories of knowledge). Not to put original research in, but to review what is there, provide missing citations, expand on the science behind an idea etceteras. What badge do you get on your sleeve (that academics recognize) for doing that work instead of chasing a grant?
One of the presenters had a system he'd cobbled together by abusing Delicious in horrible ways to try to turn it into a "cloud database". It required each user of his system to learn a series of obscure and broken ways of directly inserting data into Delicious. It was explained to me that the point of all this was that it could be done without a programmer. I couldn't help but think "but you could hire a competent programmer for a couple of hours to do that and you would have a real database where the people supporting you weren't likely to just 'clean out' your data as junk at any moment". If you're an academic, ask the university or the comp-sci department to give you a few hours of time to set up your application. Interestingly, the presenter even suffered a "cloud loss" and presented on it, with a strong message of "own your own data", but didn't seem to get that abusing someone else's system to store your data wasn't really following that advice.
Chatted with a couple of people, a few students interested in AI, a few people I knew already. All in all a pleasant afternoon, but I can't say I came out with any great ideas for how to make the world easier for scientists other than maybe "make the grid-enabled IPython/numpy really easy to use". Maybe provide some "generic" services so that people don't need to use ridiculous hacks on delicious to store key-value pairs (e.g. make one of the json data-stores available as a bald utility that you could sign up to).
Pingbacks are closed.