The Twitterverse should be a sea of fakes, but it isn’t.
Image: Karunakar Rayker/Flickr
A couple of nights ago I posted something on Twitter about the show House of Cards, which I don't like very much because of its really fake and annoying sense of cynicism (millennial cynicism? internet cynicism?). Immediately, the tweet received two favorites from @FLOTUSunderwood and @POTUS_Underwood—"Claire Underwood" and "Frank Underwood," respectively. And each account, as of Sunday night, sported around 5,000 followers with a couple hundred tweets and many thousands of favorites apiece. Bots, or at least semi-bots.
I have no idea what the actual breakdown is between fake accounts (bots) and IRL humans typing on keyboards and touch-screens. Twitter doesn't either really, but it did offer one noteworthy figure in its 2014 SEC filing: about 11 percent of Twitter users interact with the service via its API. This just means that 11 percent of accounts are not using the service's front-ends (website, app, TweetDeck, etc.), and are instead posting and whatnot via a script. It's possible to use Twitter as an IRL human via the API (I do, sometimes, just because I like looking at a Python command line more than a website) but it's not so unreasonable to assume that, indeed, 11 percent of accounts are automated. The Underwoods are, surely, the faces of some favoriting script.
Running bots is very easy to do. It's "script kiddie" business that can involve as little as uncommenting a few lines in a prefab downloaded script. You or I could make a bot in a few minutes with all of those same characteristics: searching for keywords ("house of cards") and favoriting those posts. When people pay money for Twitter followers, what they're really paying for is for someone to do just this, or this plus some other variation: keyword searching and following, following likely follow-backs, and rinse and repeat. Note that "rinse" here is literal: the key step in any automated follower acquisition scheme is the regular dumping of followed accounts that don't follow-back.
Again: your new bot could have acquired a few dozen new followers just in the time it took you to read that. Botting Twitter is effortless. Which begs an unlikely question: If creating bots—even robust bots—is so effortless, why just 11 percent?
Bear with me for the next paragraph or so.
There is an internet-popular theory in cosmology that suggests that our universe is a simulation. The simulation hypothesis comes courtesy of the Swedish philosopher Nick Bostrom and it basically rests on one variable or question, which is whether or not simulating a universe is possible at all. And if it is the case that a universe is something that can be simulated, then this universe is likely simulated. More than that, it's overwhelmingly likely we live in a simulation.
In Bostrom's words:
A technologically mature 'posthuman' civilization would have enormous computing power.
Based on this empirical fact, the simulation argument shows that at least one of the following propositions is true:
The fraction of human-level civilizations that reach a posthuman stage is very close to zero;
The fraction of posthuman civilizations that are interested in running ancestor-simulations is very close to zero;
The fraction of all people with our kind of experiences that are living in a simulation is very close to one.
If (1) is true, then we will almost certainly go extinct before reaching posthumanity. If (2) is true, then there must be a strong convergence among the courses of advanced civilizations so that virtually none contains any relatively wealthy individuals who desire to run ancestor-simulations and are free to do so. If (3) is true, then we almost certainly live in a simulation. In the dark forest of our current ignorance, it seems sensible to apportion one's credence roughly evenly between (1), (2), and (3).
Unless we are now living in a simulation, our descendants will almost certainly never run an ancestor-simulation.
So, if it's possible to simulate a universe, then it is, relative to making a universe the old-fashioned away, also very easy to simulate a universe. If the capability exists, then the proportion of universes and universe histories that are simulated (simply "switched on") compared to those that evolved over many billions of years is surely absurd.
It would take a mere one universe history to produce a sufficiently advanced civilization, which would then go on to produce an infinite number of simulated universes. Thus, it's more likely that this here is a simulated universe. A lot of people think this is a really clever and even scientific idea.
But you can see where this is going with respect to Twitter. A real Twitter account takes effort, quite a lot even to achieve an abundance of followers and to develop a brand. But it's nothing to do all of that with a bit of code in no time at all and with no effort at all. The absence of more bots might be attributed to the second Bostrom possibility (lack of interest), but it shouldn't take many people interested and capable of scripting to achieve a simulated Twitter universe, by Bostrom's logic. Should it?
Just a couple would do it, but we'd also have to assume that a truly simulated Twitter account would have to be capable of creating its own bots, which is indeed a capability, assuming the bots don't all have to be good, interesting, or particularly distinct. They don't, and neither do simulated universes have to be interesting or even complete.
I think this line of thought points mostly to the failure of the simulated universe argument on Bostrom's second point: interest. If you were to be a divine techno-creator, wielding new universes and Twitterbots like they were grains of sand in your algorithmic dump truck, why bother? What's so interesting about interacting bots and simulated universes, even if they appear to be highly autonomous? Maybe we can conclude, finally, that it's not much at all.
Then again, the Twitterverse is young and it itself is a creation, so maybe we should be talking about Twitters rather than Twitter bots. Anyhow, there's your stoned line of reasoning for the week.