On Twitter Bots and the Presence of Disinformation

On the latest episode of Winning Slowly, I made a statement that Chris took exception to but didn’t have time to mention in the podcast. So he wrote a blog post on it, which I excerpt here:

Early on in the episode, I noted that people’s feeling of decline is itself a kind of actual decline, and Stephen disagreed:

I don’t think that’s necessarily even true: because we Twitter bots that make things seem true, and… there’s literally not even anybody making that idea. Twitterbots picked it up out of the air and made it a thing…

On air, I chose to dig into some of the meatier kinds of decline which are harder to argue with. Here, however, I do want to note my disagreement with the claim as Stephen made it on air. Twitter bots don’t make things up “out of the air”! 

-Chris Krycho, “Twitter Bots and ‘Decline’

Chris is right. Twitter bots don’t make things up out of the air. Technology does not have the agency to make its own decisions, artificial general intelligence-style. However, they do a certain type of work that is very nearly making things up out of the air, which was what I (clumsily) was getting at. They are programmed by bad actors to take views that may be held by very small numbers of people (which I ineffectively stated as “not anybody”) and promote those views far beyond the reach of the very small numbers of people who hold that view. In this way, I would argue that a better revision of my statement would have been:

Twitter bots are trained to pick certain things up and promote them far beyond the original proportion of people who actually believe that thing so that it seems that many people believe something that very few people actually believe (comparatively).”

-A more articulate theoretical version of me.

Taking fringe viewpoints and mainstreaming them is not the only work that Twitter bots do; they also amplify real viewpoints and subtly twist them to different ends (among many other things). So Twitter bots are diverse, and my statement that they pick things out of the air is not only inaccurate in its technical regards, but is not comprehensive of what they actually are capable of/used for.

The post Chris wrote in relation to what I actually said is accurate, though; people do attribute agency to computers far too often. Even using the language of agency as a shorthand gives computers too much credit for what they can and can’t complete when given a task by a person (or another computer or bot; but that’s a post for another day).

In context of my argument, I was trying to say that people’s perceptions of decline must not take into account the disinformation spread by Twitter bots. Because disinformation exists, part of what we perceive is not true or real, and should not be counted as part of reality in terms of how we perceive the world. However, I would put further nuance on that point now that I’ve thought about it more.

There’s another sense in which the presence of disinformation is real and that presence itself can contribute to a sense of decline. Discussing the presence of disinformation as a factor that contributes to a perceived sense of decline counts disinformation as “real.” It is “real” in that it is an actual factor contributing to a perceived sense of decline, despite its content being untrue. Things that are not true, do not exist, or never happened should not worry you*; their nonexistence cannot affect you in a material sense. But those things being “untrue” does not necessarily mean that the disinformation does not exist–even if it should not exist. And the presence of disinformation can contribute to a perceived sense of decline (and perhaps rightfully so); the actual disinformation can contribute to a perceived sense of decline (but I am saying that this should not be so, and we should fight against this tendency).

Ultimately I don’t think Twitter bots make things up; they only do what people program them to do. But the act of programming them and the presence of the disinformation they sometimes spread have effects that I was not fully mentioning at the time.

*not that this stops me or anyone from worrying about the future, but this is my contention in a philosophical mode. It is a mere contention, and there are arguments against it.