Autonomy, packet size, friction, fanout, and velocity

Nostalgia is a dangerous drug and it’s always risky to wallow in it. So those of us who fondly remember the early blogosphere, and now want to draw parallels to the fediverse, should do so carefully. But we do want to learn from history.

Here’s one way to compare five generations of social software along the five dimensions named in the title of this post.

              Autonomy    Packet Size   Friction    Fanout    Velocity
Usenet        medium      high          medium      medium    low

Blogosphere   high        high          high        low       low

Facebook      low         high          low         medium    high

Twitter       low         low           low         high      high

Fediverse     high        medium        high        medium    medium

These are squishy categories, but I think they surface key distinctions. Many of us who were active in the blogosphere of the early 2000s enjoyed a high level of autonomy. Our RSS readers were our Internet dashboards. We loaded them with a curated mix of official and individual voices. There were no limits on the size of packets exchanged in this network. You could write one short paragraph or a 10,000-word essay. Networking wasn’t frictionless because blog posts did mostly feel like essays, and because comments didn’t yet exist. To comment on my blog post you’d write your own blog post linking to it.

That friction limited the degree to which a post would fan out through the network, and the velocity of its propagation. The architecture of high friction, low fanout, and low velocity was a winning combination for a while. In that environment I felt connected but not over-connected, informed but not overloaded.

Twitter flipped things around completely. It wasn’t just the loss of autonomy as ads and algos took over. With packets capped at 120 characters, and tweets potentially seen immediately by everyone, friction went nearly to zero. The architecture of low friction created an addictive experience and enabled powerful effects. But it wasn’t conducive to healthy discourse.

The fediverse can, perhaps, strike a balance. Humans didn’t evolve to thrive in frictionless social networks with high fanout and velocity, and arguably we shouldn’t. We did evolve in networks governed by Dunbar’s number, and our online networks should respect that limit. We need less friction within communities of knowledge and practice, more friction between them. We want messages to fan out pervasively and rapidly within communities, but less so between them.

We’re at an extraordinary inflection point right now. Will the fediverse enable us to strike the right balance? I think it has the right architectural ingredients to land where I’ve (speculatively) placed it in that table. High autonomy. As little friction as necessary, but not too little. As much fanout and velocity as necessary, but not too much. Nobody knows how things will turn out, predictions are futile, behavior is emergent, but I am on the edge of my seat watching this all unfold.


Posted in .

4 thoughts on “Autonomy, packet size, friction, fanout, and velocity

  1. Zooming in on what you wrote here, “…less friction within communities of knowledge and practice, more friction between them. We want messages to fan out pervasively and rapidly within communities, but less so between them.” Seems like the values in some of the columns would move a bit between being in a small instance with a clear topic/focus or in any case a lot of conversation among themselves (e.g., just found out about ), vs. a larger general once such as

    There may be even more nuanced ways that different kinds as well as sizes of instances affect those values. Anyway great post, it’s stimulating further thinking even if I’m not going to follow it down any of those potential rabbit holes right this moment.

Leave a Reply