Published on 2017, Jun 8
I wrote about an event I observed on Twitter and how it pertains to popularity effects I’ve seen in real life and especially amplified on social media.
On Twitter, I follow several accounts which are rather inter-connected in their topics and interactions, such as @SwiftOnSecurity, @AwfulyPrideful, and @MalwareTechBlog. From here on out, I will be referring to these users as, respectively, Tay with “they” pronouns, Amanda with “she”, and MalwareTech with “he”. (If you are one of these people and I made the wrong choice, please let me know and I’ll correct.)
Tay is an enormously popular account, and I will admit I absolutely followed them first by a significant lead. Through their feed, I was introduced to Amanda and MalwareTech, and as I enjoyed what they had to offer, I followed them as well.
Amanda has already written some excellent backstory and commentary on the events prompting this particular post, so I’m not going to rehash the baseline. Go read her piece and get back to me; I’m just providing third-party thoughts tangential to the main event.
If you’ve ever said something to little recognition only to have a more popular person say that same thing and get more recognition, you can sympathize with Amanda’s professed “pettiness” and “saltiness” (actual quotes not scare quotes).
Amanda’s not upset with MalwareTech about either instance, since as far as anyone knows MalwareTech didn’t plagiarize her work but rather seems to be an instance of great minds thinking alike. There’s been a little bit of misunderstanding about that and I don’t want to add to it myself, so, kindly presume that everyone in the story is on friendly terms.
Anyway, on to the actual point I have to make with this piece.
Social media in general has some serious flaws with the way social interactions and conversations are modeled. Each network is broken in a different way, but as this piece is prompted by events occurring on and somewhat particular to Twitter I shall focus only on it.
In pretty much every medium, the first person to make a comment is considered more important than anyone else who makes that comment, and the relative importance of any particular person making a point decreases as time from the original remark occurs. But Twitter’s timeline, like that of other live feeds across the internet, is Newest First. All else being equal, if Amanda and MalwareTech post before I start scrolling, I will see the second post first and the first, second.
This is probably not what Jesus meant by “the first shall be last and the last shall be first.”
As a reader, I will come across MalwareTech’s tweet first, and have a genuine reaction, possibly including liking/retweeting, to it first. By the time I see Amanda’s, I will have already reacted, and will probably not be moved by hers. If the two posted independently, this is merely irritating; if MalwareTech had seen Amanda’s and decided to repost it himself (which is not the case; I just only have the one example scenario) then it is downright unfair and rude.
MalwareTech has the bigger audience by almost two orders of magnitude; it stands to reason that his tweets would reach a larger audience. This is fine, on its own. Twitter, however, quickly finds ways to make it less fine.
I am an audience member for both MalwareTech and Amanda, so from my perspective it doesn’t matter which of them is more popular; I regard both with equal fervor as I follow both directly. Twitter, however, shuffles my feed so that more popular posts appear before less popular, and this is judged both by impressions on each post, but also by the audience count of the poster. MalwareTech is approaching 100k followers, Amanda just cleared 600; clearly MalwareTech has “better” content according to Twitter and his cuts in line. This is terrific behavior… when your platform is trying to min-max people’s time and not bother to require that they invest attention without a clear indication of reward. Good for ad revenue, bad for actual conversation. If it’s not apparent what is and isn’t Twitter’s service and product yet, hopefully the picture is growing clearer.
This setting can be disabled, but it’s apparently on by default. I certainly hadn’t turned it on manually, as I dislike this behavior, but it was on for me when I was told the switch existed.
If you’re in a conversation, which do you value more? Someone in the group responding to you, or someone grabbing a bullhorn and speaking at, not to, the crowd?
Most people would value the former, and consider the latter less cool. Twitter has the opposite priorities: replies are limited to those mentioned in the thread already, and the followers of the speaker.
Amanda made her joke in reply to Tay; I might or might not have seen it on my own merit, though if Tay with their massive popularity had reacted to it, that would have helped. Replies to Tay do a lot better when Tay likes, responds, or retweets them, than when they do not visibly engage.
MalwareTech retweeted an article directly with his commentary; Amanda provided it as a reply to Tay’s tweet about it. As MalwareTech’s was a top-level post, it got primacy of audience; as Amanda’s was a child item, it did not.
Granted, it is basically impossible for Twitter’s algorithms to note this, and I don’t think anybody is demanding a change. But that is still a noteworthy effect of the platform.
I don’t have one, really. I was a bystander to the example provided, and only even got pulled in because I happened to be in a position to observe, and noted that to Amanda.
I do think that the effects social media are having on the general paradigm of social interaction are interesting and worth talking about. I’m not going to condemn the internet in general or social media in particular as some evil of technology without which the human condition was better – I greatly appreciate that these technologies let me stay in touch with my friends halfway across the country, and get information I enjoy from an InfoSec person who uses the guise of a pop star to flavor their message, and also write this article in the event that you or anyone reads it. But there is a spectrum between Universal Good and Universal Evil, and social media is not without its flaws. The upheaval of social norms to which humans are long accustomed is, at the very least chaotic and disturbing, even if the end result may be better in the long run, and the admixture of conversation and capital certainly leads to some extremely twisted events in the pursuit of wealth, be it monetary or influential.
Just something to think about as we try to adjust the momentum of collective consciousness to new technologies, I guess.