I’m a librarian, so Aleks Krotoski’s question yesterday rings a bell for me. Libraries are about arranging things physically to facilitate serendipity. There are some advantages to that, like putting everything by and about William Shakespeare in the same place so you can (barely, in a small library) see all of it at once and browse through it quickly. But there are disadvantages, like the way our books on political science are unpredictably scattered between libraries on opposite ends of campus. They sit together on our virtual shelf, the library catalog, but in space they are far apart.
The web facilitates serendipity on a different scale. About ten minutes ago I was reading a blog on pens, then wandered into Tumblr’s tech tag and found a column I’d never read before published by a famous newspaper on the opposite end of an ocean. I never read The Guardian before the web, because I never had access to a library with a subscription; but it’s a trusted source for me now. I agree, you can use the web to filter what you see and ensure you only see what suits you. Or you can use it to open up a world unimaginably more diverse. Very easily. You could still filter in the analog world, though. It was just much, much harder and more expensive to do the other thing. Is the web the ultimate serendipity engine? Yes, if you do it right.