There is an inherent tension in social software between content discovery and the quality of conversation around that content. Group conversations get worse as groups grow, and groups grow as group discovery improves — if it’s easier to find something, more people will find it. Therefore, the easier time I have finding good conversations, the less likely those conversations are to be any good (e.g. Reddit’s front page vs. subreddits). Paradoxes should be named, so let me know if you have any good ideas.

Let’s look at Twitter through this lens. Twitter began as a space for conversation — a messaging platform. It exhibited characteristics of a “many-to-many” network. Anyone could publish, anyone could follow anyone else, and “discovery” in this context meant discovering people to follow, not content to consume.

Over the last few years, we’ve seen Twitter evolve its focus towards the discovery of content — hiding @mentions from user streams, the #NASCAR page, the Summify acquisition, the new Twitter email, the Discover tab. Today, it looks more like a broadcast medium than a distributed social network. Large groups of users (consumers, really) follow a small number of very large brands. Some of those brands are people (Bieber) and some of those brands are publishers (@cnnbrk). Lots of talking, very little listening.

Why is it in Twitter’s interest to focus their users around a relatively small number of mega-publishers?

First, Twitter needs to reach late adopters with a product that late adopters can understand. It’s much easier to bring people onboard to “a real-time feed of news links from publishers and celebrities that you’ve heard of” than it is to explain “a distributed messaging platform where you follow friends, some people you don’t know, some celebrities, and some brands…where you make lunch plans, share what you had for breakfast, and post your favorite links of the day…where sometimes it’s a chat room and sometimes it’s the nightly news.” Try explaining Twitter to your parents and see what works better. On a unique active user basis, evolving into a more traditional broadcast medium will be a boon for Twitter.

Second, brands (you know, the $$$) don’t know how to join small group conversations. They do, however, know how to shout at large groups of passive media consumers. If Twitter looks more like a broadcast product, then brands will have an easier time fitting Twitter ads into their campaigns (and budgets).

As with any piece of social software, as Twitter evolves from a space for conversation into a space for discovery — prioritizing features that support a one-to-many model at the expense of the many-to-many model — we will see its value as a conversational platform erode.

For lack of a better term, you might call this the “broadcast-ification” of the major social media platforms. I’m picking on Twitter, but it’s happening in different ways across the industry — see LinkedIn Today or Facebook’s asymmetric “Subscribe” feature. These platforms are all evolving towards a more traditional broadcast media model, because it’s more palatable to late adopters and because that’s the environment in which brands know how to communicate and, more importantly, spend.

So 2013 will bring two things: more ad revenue for the major social media platforms, and a massive opportunity for upstarts like Branch, Reddit, Digg (hopefully), and a company or two that doesn’t exist yet, to create spaces where small groups can engage in high quality conversations.