Sleep is not coming, so here's my condensed ETCon thoughts:

1) Ambiguity Is A Basic Human Need. Computers Are Shite At Ambiguity.

The talks by Matt Webb and Danah Boyd were two highlights for me. They hammered home the point that social interaction is lubricated by ambiguity. Matt described people needing the option of plausible deniabilty in order for small groups to thrive. White lies are important.

Danah talked about people's need for several, sometimes independent contexts - you might think your boss is awful, but you really shouldn't tell her this to her face. But then you couldn't pretend otherwise to a colleague. Likewise, if you're a teacher, you really don't want your pupils raiding Friendster to see who you're dating or with whom you go clubbing.

Social software - indeed computers full stop - reduce ambiguities to a minimum. By design. Big Problem.

2) Technorati as a feedback tool for journalism.

Say BBC News Online parsed the Technorati API looking for incoming links from the blogsphere to that week's stories, then published those links and synopses. You need to think the precise mechanics through very carefully - the links and synopses would need heavy caveats, and would have to be on a page one-click removed from any BBC journalism itself. Also, why not let BBC News journos sign up to a daily email listing comments about their stories in the Blogsphere?

3) Technorati is an API'd alternative to Pagerank for Site-Search

David Sifry's talk on Technorati got me thinking about their free-to-not-for-profit API. In some ways it's a very cheap way to get something not a million miles away from Google's hugely closed pagerank data if one was looking to derive relevancy from incoming links and associated link text. Am toying with how we might use it for a forthcoming project for which incoming links are going to a significant indicator relevance.