The Robo-proletariat

In a continuing cycle where Ben Abraham writes something, Daniel Joseph responds, and I comment on what they both wrote, here I am doing it again.

Ben wrote a post titled “The End-Game of Labour Automation Meets Social Media” in which he analyzes the work of bots in creating social relations on social media websites. He writes:

Let’s look at an example: the ‘About Birds‘ Facebook page has 9-likes. One of them is me. Who else has liked this clearly spammerific etsy-esque store attempting valiantly to sell bird-decorated products? Actually ‘About Birds’ is just one of a suite of similar pages – About Monkeys, About Dolphins, etc – and some of the people who interact with these pages are clearly, well, bots or at best sock-puppets, and they appear to constitute the majority of the ‘likes’ on the page.

So the “work” of social media as it is framed by Facebook, which is liking, sharing, and generally hawking on a page, is being performed by, at best, a being that is untouched by a human for 75% of its “working day.” On the other end, it could just be a bot that hangs out, creates social relationships, and likes pages.

Daniel responds and adds to Ben’s post. The main takeaway from Daniel’s post is this:

The marketers know there are bots out there messing up the accuracy of the data, but this only matters if capitalists see the bots as fundamentally altering the value of the data. If the capitalists do, either facebook will need to invest heavily in anti-bot programs, or have their business, and its value extracting ability, melt away. What we see here is the ongoing struggle between workers (us) and capital – between the real people labouring to produce value and the vampires. When capital does hit this limit, because of spammers or activists like Darius (random bots 4 lyfe) [f]or Ben it will constitute a moment of decomposition.

We live currently in a field of production that is shot through with the nonhuman. Of course, human lives are constantly intimate with the nonhuman, but this if different; there is a world where bots, and their connective/informative capacities, are actually doing labor. They are connecting. They are making things work together. They are generating information.

I take a different turn than Daniel does above. I don’t think that we will hit a limit, a moment where advertisers will say “damn it, we have to get away from these bots!” The bots, as I see it, are crucial to keeping the entire myth of capitalist production, of the virtual of capitalism as Bifo would have it, operating. The bots function as immaterial, perfect laborers–they “like” and link, connecting to one another, and more importantly, to humans.

They make themselves, like Latour’s ARAMIS, more real with each human who accepts a blind friend request. Photos lifted from imgur become a personal history. And so, like many of us, the bot labors not in a specific way. They labor through pure connection, and I’m wary of reducing this to the interruption of data value–instead, these bots become a kind of gold standard for value. It isn’t about eliminating the bots from the system, but eliminating the ahistorical, the properly “nonliving” bots, from your data set. In this way, the more “living” of the bots, those sharing pictures and liking pages and having a thousand participating twentysomething friends that maybe knew them in high school, those will be included.

Because they’re producing, they’re alive. And because they’re alive, they count.

“like”-ing as living


This entry was posted in General Features, Theory and tagged , , , , . Bookmark the permalink.

3 Responses to The Robo-proletariat

  1. Ben Abraham says:

    Interesting take on it, and I really appreciate the Latourian turn to “counting”. That’s really important, I think. But I’m also not sold on this aspect:

    “Photos lifted from imgur become a personal history. And so, like many of us, the bot labors not in a specific way. They labor through pure connection, and I’m wary of reducing this to the interruption of data value–instead, these bots become a kind of gold standard for value.”

    So, this might just be a sign of slightly different interests, but it seems to me the more pressing question is lurking behind the “personal history”. The only reason any bot can appropriate (or approximate) a “personal history” is because it is *assumed* that a personal history is something that a “real person” has, or would do. it’s like there is an idea, or a dream or fiction, of a “real personal history” – and I’m absolutely sure I can’t put my finger on what that constitutes, but maybe that’s part of the point… it’s a dream so it’s easily recognizable by anyone, and it needs to be or else we end up aliens to ourselves and each other, spinning in space alone.

    I guess what I’m trying to say is that for me, the path I’m going down at the moment is about recognition – it’s all about whether or not someone “recognises” and extends recognition to the “reality” of the human, which I think we do a LOT – we don’t see “a facebook profile” we see “Cameron Kunzelman”, and bots make this a bit more apparent b/c there is no “[name]” behind the profile, at least not in the normal sense. Like the Turing test – that de facto standard – which has been argued (by Bogost!) as really being about pretending and imitation, I think that’s really important. Not even Turing could give a non-relative account of what intelligence *is*… I don’t think?

  2. Saelan says:

    I don’t see this as being a matter of whether bots or alive or producing or not, but a matter of how to extract capital. Bots might be able to add value, but unless they’re Darius’ random shopper, bots don’t spend, and if, as Dan suggests, bots and spam could proliferate to the extent that they overwhelm the human content of the network, then that seems like it might be a real problem for value extraction.

  3. Ben Abraham says:

    Also, I really recc’ this piece by Peter Lewis who commented on my initial FB post. He wrote about this and mentioned some idea I hadn’t heard before:

Comments are closed.