There have been a number of posts on Metafilter and especially Buzzmachine about how the reduced costs of communication associated with the internet have changed the job description of the media. Reporters as specialists are being overwhelmed by numerous nonspecialists who, in aggregate, make up a kind of market wisdom-- the pajamahadeen (spelling?). Control of access, content, and even reportage are going to pass over to the consumers of news; and current organizations which ignore this trend will likely get bypassed. (I just spent part of the morning looking in vain for an RSS feed for the LA Times. Since the WaPo and NYTimes have them, this means my daily news aggregator includes them, and e.g. Slate, but not the LAT. )
What's important in this trend (see especially the Metafilter comments, which resemble the spats between the Palestinian People's Liberation Army and the People's Liberation Army of Palestine in Life Of Brian ) is that decreased costs define the sea change. Items which still cost are less affected. For example, it costs very little to be a blogger (caffeine expenses nothwithstanding) *BUT* is still costs quite a lot to maintain the internet backbone. (And Google runs blogspot with some anticipation of profit as well). So the costs of some kinds of dissemination have fallen through the floor, but, to stretch the metaphor, the house still has the same four walls.
I remain very interested in how these changes might influence the way science is done and disseminated. I operate in two distinct worlds as scientist blogger. On the blogger side, reduced information costs make it very cheap for me to learn quickly (at the dilettante level) about a big variety of things. I'm loving it, but don't take the product so very seriously. On the scientist side, what has not changed is thinking, messy results, and general banging of my forehead against rather pricey walls. This is Coturnix's repeated admonishment that good science is muddled out in pencil on paper, before beginning the technical side.
(Nevertheless, connectivity has made deep inroads. I have not been to the paper library at our institute in my entire 3 year stay, because I download everything.)
Back40 at Muck and Mystery gives a few hints about what a distributed science might look like. Diverse groups of problem solvers can outperform groups of high-ability problem solvers, within parameters. The distinguishing feature of this "skunkworks" model, in my opinion, is its (dare I say it?) collectivization of both effort and credit. And, within parameters, this is going to be the better way to do things. I think for some very hairy scientific problems, the pajamahadeen, distributed model is going to be the way to go. I'd better sit down with Ayn Rand and have a talk about this.
UPDATE: Some more thoughts: At Tech Central Station, Glenn Reynolds is a known blog triumphalist, but in any case he's saying similar things about changes in the media. I thought a bit more about the science side. It seems that institutional science is actually quite dynamic, with certain data-generating procedures very rapidly outsourced as soon as they become routine. In my own time, oligo synthesis and sequencing have become entirely outsourced; and cdna cloning is not far behind. That is, scientists' job descriptions migrate pretty steadily toward the expensive, the once-off, the technically demanding. This occurs for many reasons, but it may mean science is a different sort of beast entirely than the newspaper industry. (See Back40 on Celera's human genome coup however. There are disruptive technologies in science as well.)
YET ANOTHER UPDATE: Teambuilding at New Scientist . In their case a pecking order is built in.