Unexpected expertise
If enough people are in a conversation, one of them will be an expert. The larger the crowd, the more unexpected will be the expertise contained within it.
Of course, "larger" in this case may mean thousands, or tens of thousands. And, to uncover really obscure expertise, you may need millions of people. Of course that also means that you'll need a social environment where obscure expertise can rise to the top. But that's supposed to be impossible: Conversation doesn't scale, we were told.
We were told wrong.
We're now seeing social media in which thousands may participate, and millions may audit with the option of jumping in. For example, a few months ago there was a thread at Reddit.com that was supposed to be discussing the six months a book author spent riding trains. The highest-ranked question was about what the worse railway station in America is. The answer: East St. Louis. Of course, there isn't a single right answer to such a question, but the Reddit thread drew a lot of comments confirming that there are many safer places in which to loiter.
That's a type of local knowledge, but there's more on display in this East St. Louis thread. It meanders into a discussion of knife fights, which in turn leads to some EMTs and hospital emergency room folks talking about the types of damage done by various types of knife wounds. And this is in a thread about riding trains. Get enough people into a room for one original reason, and you will discover experts in fields completely unrelated to that original reason.
This makes sense once you see it happening, but it wasn't a predictable outcome of being able to hold conversations at scale.
Sort of serendipity
It's also 90 percent wonderful. The 10 percent is that the experts may be full of baloney while being able to talk a good game. It happens, but in a well-constructed conversational space, the ability to interact and contradict should provide some measure of defense against the incoming horse-hockey pucks.
The 90 percent wonderful comes from the ability of conversations to be extended and deepened by people who do know what they're talking about. This is the very best sort of serendipity: knowledgeable but unexpected outbursts of learning.
It is the sort of occurrence we want to optimize our social systems for. We want to enable particularly helpful spontaneous expressions of expertise to be prominently displayed; it does no good if they're there but are 4,000 comments in. We want these experts to be rewarded as well, although often being prominently displayed is reward enough.
We also want to make sure that there is enough diversity in the readership that there can be unexpected outbreaks of expertise. For this I find it's useful to keep in mind the words of Bernice Johnson Reagon, the founder of Sweet Honey in the Rock and a social activist: "If you're in a coalition and you're comfortable, you know it's not a broad enough coalition." She was talking about political coalitions, but it's also true of any conversation at scale: If the conversation isn't making you a bit upset, then you don't have enough different ?types of people talking. Because you're looking for expertise that may not be related to the opening topic, you want a lot of different sorts of people involved.