True serendipity is when something batshit crazy just comes out of nowhere when you are not looking for it at all.

The skill is to be able to recognize it when it appears.

“There is a crack in everything
That's how the light gets in. ”

Anthem Leonard Cohen

well i'd say basically for all kinds of content there's two axes:

- being able to find exactly what you know you're looking for (fast, easily)
- someone/thing recommending stuff you didn't know you wanted (with
minimum wasted effort)

Looking beyond the social aspects, the international limitations, the
device availability, service inertia and the prices (hmm…), and
ignoring that the services above can be complements to each other,
discovery and serendipity (finding new music we might like and love)
is a big part of how and why we choose each service. Each uses their
own set of filters with different uses of algorithms and people. For
many of us, the choiceon how we allocate our time and money comes down
to which delivers to us the best music. In this case, delivering the
right level of serendipity becomes a very practical business question.

And thus, your reformation is spot-on.

Your last point is the best, though: yes, it's possible to experience serendipity without inefficiency. But the question then becomes are you getting enough serendipity? The argument is similar to the discussion we had about privacy and location-based services with Alan: in the same way that people might “overvalue their privacy and undervalue the benefits of broadcasting more info publicly (both on an individual and aggregate level)”, we could also be undervaluing the benefits of serendipity (i.e. continuous partial attention, yada yada).

Maybe.

The point about reputation score and velocity is perfect, and I think more metrics like that will help us figure out how to balance our attention and time across showing up and participating online and offline. This is why things like twitter lists, influence scoring (e.g. @tunkrank, whuffie bank), postrank, backtype, Radian 6 etc all matter: because they are trying to figure out all these little decisions we make in how we balance our time and attention and condense them into numbers that help us make informed decisions.

1. clicking on links in blog posts/articles I'm reading (pre-Twitter era)
2. surfing the “stream of stuff from people I'm subscribed to (Twitter era, also incl FB newsfeed of course)

The primary function here is: trusting the sources enough, in aggregate, to allow them to waste a bit of your time as an investment. The return is finding something that you didn't know you needed/wanted.

So I think that this is getting at the same “attention economy”/info mgmt issues that have been discussed elsewhere, but from a different and useful perspective (“serendipity”).

That is to say: I think it's possible to have a stream of 100% serendipitously awesome stuff I wanna click on, with 0% “noise”. Or: serendipity doesn't necessarily have to carry the cost of inefficiency. A combination of a reputation score (quality over time) and velocity (what content do most people like right now) would probably solve this.

Just putting this out there.