Wednesday, April 22, 2009

Cloudthink

The hype around "crowdsourcing" (a cool term coined by an intelligent but hype-y author) makes me wonder how it is that the hype factory is unaware of its own groupthink. I am going to coin a new term: cloudthink.

Cloudthink is what we collectively get when we have Amazon, or any other taxonomists including the group of Flickr users, do all out thinking and categorizing for us. We think like a big massy gassy cloud. The technology is neutral - it displays our patterns. It shows us as we are: the greek choir, the mob who sees Antigone's civil disobedience as bad for the state. Group categorizing, whether by closed group of highly educated taxonomists, or an open group of global citizens, has no inherent good, and is certainly not democratic. Its just a methodology. The lighter the governance or adherence to "universal" principles, the more likely the group is akin to the Greek choir; seeking balance and justice, but sometimes demanding too much in the heat of emotion.

Evidence pieces in point: Ushahidi in Kenya crisis, Amazon in non-crisis. Both tools in which group categorization of information is heavily in play.

One one hand, you have Ushahidi, this amazing project that aims to use distributed technologies to collect (or "crowdsource") incident reports/news for democratic aims in the Horn of Africa region. During recent Kenyan post-election crisis, the ability to transmit on-the-ground reports by eyewitnesses and make it available to each other and the outside world enabled a citizen reporting capability that really walks the line between journalism and activism, but nonetheless seems more dedicated to truth in reporting than the average schlock on US cableboxes.

Absolutely amazing and can't express enough admiration. Collective categorizing and tagging allows for easy fast mass collection of data. One of the really exciting possibilities here is the collective filtering and and collective pattern generation- the trend over time, the peaks and the valleys. To understand the cycle of violence, in order to have advance warning. Or to use the connective tissue of the network itself to enable better reporting response if it happens there or elsewhere. But again, the tool is actually neutral. It could be used by any group to achive whatever categorization of incidents. My incident "Tutsi murder" is your incident "Hutu justice."

Second case: Amazon's #amazonfails mini non-crisis. Capitalist behemoth, online winner despite all recessionary signals. Seemingly little relation to social justice. Their taxonomists are centralized - a small group, one would assume, of specialists which categorize and tag resources (in this case, books). In contrast to Ushahidi, the technology allows users to contribute to the body of tagging, but not override the underlying taxonomy. The taxonomy is set. The tags are secondary. SO useful, so easy to find things.

The collective filtering here then, is also secondary to the main taxonomy. The effects in these two functions could both slightly repulsive in a social sense. First, the taxonomy itself is subjective and can be manipulated or modified in error. Case in point: the "gay" search results. Second, the collective filtering effect that leads to further and further polarization of thought - the recommendation engine only recommends things that are similar, not different.

In both Amazon and Ushahidi casesabove, there is no real news. However, you could say that not enough thought is being given to the balance between the wisdom of crowds and the madness of crowds. (Long overdue, I am just reading Extraordinary Popular Delusions and the Madness of Crowds.) It always seemed that Shirky and others were conflating collective participation with democracy, (an of course democracy being the shiny golden semi-religious ideal). How many articles, posts, blurbs, soundbites have I heard about the incredible wisdom of crowds? Given our problems with the big structures of our American lives -- politics, economics, etc. -- perhaps the joy at the door being blown off publicly created information (both access to and ability to contribute to these public spaces), and scores of private sources, itself was a kind of frenzy or madness.

But Ushahidi and Amazon (and flickr, and delicious, and bla bla bla) make the same oversight: conflating mass participation with actions guided by (old-skool, yes, but still grounding) universal principles like democratic activity. I might argue that collective participation can turn into democratic activity if it is guided by principle. In fact, Shirky seems to suggest this in the first days of social software. Its no surprise that we have untamed human natures. Thats why we created principles to begin with, no?

The point is that both Ushahidi and Amazon can be gamed, both by specialists on the inside, and the Thebans in the commons. Thebans collectively revealed the truth that no individual could reveal - the origin of Oedipus. But they also demand justice, and proclaim maxims that are only true unto the group in a specific point in time. I wouldn't argue for more regulation, god no. But if we are aware of our group behaviours, at least we can take action.

Otherwise, the machine is being fed by the happy mob. Cloudthink.

Thursday, April 16, 2009

keep it together or merge with the universe

So the latest brainstorm has been over access. What is access? Who decides it, by what criteria, using what logic, who controls it, who overrides it? There are some strong loci of theory and practice around communities of librarians and archivists, freedom of speech and civil rights practitioners, cryptographers and security analysts, and IT guys.

The above curveset is not quite done, but already it seems striking to relate it to the curve to the right describing learning and cognitive information processing.

Am I trying too hard to reconcile a cognitive model with policy?