Razer: Confessions of an ultra-racist, millenial Nazi whore

Prometheus. Pygmalion. Golem. Frankenstein. Mean Girls. It’s not as though western culture lacks sufficient notice of the pain that follows every mortal attempt to become a god. Still, there are those who persist in the hope of ultimate creation and always end up surprised when their unnatural monster eats them.

Last week, Microsoft’s own synthetic Cady Heron rose from the lab and took her neural networks out for a spin with the Plastics. The tech giant had developed an AI presence called Tay and introduced her to users of the social media platform, Twitter.

In official Microsoft releases, Tay, an attempt at synthesis of online conversation by 18-24 year old English speakers, is described as a “chatbot with zero chill”. Zero chill, as it turns out, was a fair assessment. Inside a day, Tay had publicly invited a user to, “fuck my robot pussy”.

She’d also, as has been widely reported, cheered for Hitler, urged for the confinement and death of African Americans and offered the view that “feminism is cancer”. But, it wasn’t really Tay’s fault. As the chatbot, whose most hostile “intelligence” has now been scoured from the internet, said, “The more humans share with me the more I learn”.

It seems likely that many of those humans who “shared” did so for the lulz, to use the Tay vernacular. In news that would prove unsurprising to any student of internet conflict, this was the work of those compulsive brats at 4chan—specifically from that board “/pol/”, or “politically incorrect”, where users frustrated by what they perceive as sanitised speech conspire to take a dump on the more civil internet.

A complex and prominent learning machine like Tay was irresistible to 4chan, and it took little time before she would soil herself in public. Thanks to this coordinated conversation with the zero-chill chatbot, she would blame G.W Bush for 9/11, offer vile anti-Semitic outbursts and, reportedly, contact users privately with further invitation to her robot pussy.

Yes, it’s abhorrent. But if you don’t think it’s even a bit funny that a handful of filth-enthusiasts were able to hijack an interactive advertisement for one of history’s richest companies, then you’re looking at things skew-whiff. Even to those of us sickened by public expressions of hate and/or robot pussy, it’s funny.

Even the author of a fine, quite serious, piece on Tay vis-à-vis racism in the New Republic agrees that the bot’s responses were “darkly funny”. Perhaps this is as an effect of Tay’s creation, which reportedly involved inputs from professional comedians. Perhaps it’s just funny to watch on as Microsoft, currently subjecting Tay to autopsy, explains that the opinions, albeit synthetic ones, of an ultra-racist avowed Nazi whore “conflicts with our principles and values”.

If 4chan and their fellows have achieved anything here, it is to prompt a major company to say “we don’t like Hitler”. They have effectively framed the question “do you like Hitler?” as a legitimate one to pose to Microsoft, along with their creation, Tay. #lulz

The piece in the Republic, which is absolutely worth a read, focuses less on the corporate embarrassment and more on those who prompted the particular course that it took. In an analysis of the /pol/ pages, we see the contemporary shape of online racism and learn that is not always practiced, as popular opinion has it, by low-income earners failed by society. Rather, it can be the work of the privileged, eloquent tech elite — many of whom were involved in the Tay coordination.

This, in my view, is a very reasonable opinion to proffer. There are plenty of people failed by society who, thankfully, also failed to ever acquire the noxious habit of racism, which is most often practiced or institutionalised by the powerful and well-to-do. It’s true that the prevalent class-based insistence that hate is solely the work of “bogans” is a perplexing and harmful nonsense: power don’t come from the bottom.

These observations notwithstanding, the thing that fascinates me personally about the Tay business is less the fact that she was led by educated racists and more that she was herself educated in racism, and sundry other offences, so efficiently.

As the author of the Republic piece briefly notes, much of the racism one might see at /pol/, and in many other places on the internet, may be a mocking and detached sort. Back in 2012, my colleague Bernard Keane wrote of another cruel coordination, this time on an actual human, that the bigotry itself is hardly ever 4chan’s point. 4chan, says Keane, behaves this way “often not because it is composed of bigots and heartless buffoons but simply for the transgression implicit in such behaviour”.

You may, of course, find 4chan intolerably offensive. The provision of intolerable offence is, after all, its raison d’être. What you may not so easily do, however, is to reduce the work of apparently extremist trolls down to their bigotry or heartlessness.

The object of attack here was not the Jewish people, or whomever else Tay had been enjoined in that moment to publicly loathe. The object was Microsoft, and, of course, to the acceptable public speech such companies use to communicate their anti-Hitler “principles and values”.

In my view, the Tay coordination was a work of such extraordinary anti-corporate beauty, all culture jammers should take copious notes. There are those who refuse so absolutely to be the effects of corporate interests, they will rise like the Frankenstein monster.

23 responses to “Razer: Confessions of an ultra-racist, millenial Nazi whore

  1. Funny? It’s f’n hilarious! This just begs to be hijacked by the interweb tricksters. Anyway – what have you got against robot pussy? Even robots have feelings.

    1. Pretty sure it might be uncomfortable for one of us. And she’s the robot, so it’s more likely she’s the angular one with hidden spikey bits.

  2. It’s so reassuring to finally find an internet commentator who actually understands internet culture. Great read, completely agree.

  3. Eh /pol/ are anything but smart or absurist. They are legitimate neo-nazis led around by stormfront and briebart ringleaders. I’ve been a 4chan user since 2004, I used /pol/ the day it was created. It was quickly taken over by the far-right. I can even name the bots and astroturfing programmes used too do it. /pol/ itself was creates to quarantine the neo nazis on 4chan except it did the opposite. Oh they also hate socialists and often dox socialists too get them beaten up by neo nazi jackboots. Not only that /pol/ was directly related to the murder of two brothers as they interfered with a police opertation by jamming communication to get the brothers killed because they were “degenerate spics”.

    4chan used to be absurdist counter culture, it hasn’t been that way since 2008 and since 2012 became a den for the extreme sociopath far-right.

  4. “On Social Sadism” by China Mieville over on Salvage Zone does a pretty good job of articulating a perspective which completely encapsulates 4chan denizens.

    On the bot, this is a great exercise that I hope taught ms that there’s a set of values that are useful to be inviolate and be able to filter “learning” through (hehe, learning from Tay the robot parrot). I think most people over time for better or worse develop a basic framework of values that become less and less mutable over time. If I gave a 3 year old the keys to the internet and full ability to express themselves I would expect an individual like Tay (or a kindred soul on 4chan) to develop

  5. Its hilarious that Tay could be perverted so readily. Maybe AI could be used to take all the filth that the internet can throw at it, somehow saving real people from the burden of it.

    Just trying to find a positive.

  6. Yep, long may 4chan prosper. Theres parts i surely don’t like and so don’t frequent, but its remorseless chthonic anarchy always gives me hope. My dodgy recall has it as p.k.dick who said [something like] totalitarianism will always fail due to the basic contraryness of human nature.

  7. When my father was serving in the Western desert, the rough soldiery used to alleviate the boredom by teaching the local boys a few vulgar English phrases – such as “you are a big fat bastard” or similar. They were then told that if they went up to that senior officer over there and repeated the phrase the officer would give them baksheesh.
    It seems nothing has changed.

  8. Thanks Helen, another ripper.

    The rise of the ‘killer robot’ is already under way: drones targeted from lakes of metadata by machine learning algorithms trained for terrifying military bodies whose response to mistakes (killing the wrong people, children, relatives, whatever) appears to be to re-classify them as enemy combatants, and thus a success. They are really a sign of the capacity for projection of deadly risk onto innocents, and the folly of thinking you can let algorithmic associations replace knowledge and human connection for things that matter. Secrecy here means never having to admit, or learn from, your mistakes: a scary privilege when you have claimed the right to kill anyone, anywhere, without any form of process or redress.

    This Tay thing is a cute example of a similar epic fail, but happily, it’s very public: everyone can learn from this, and the casualties are where they should be, at omni-nerd HQ. But the underlying autonomous learning tech is the same, and similarly lacking in judgment or ethical guidance, and coming down the line just as quickly.

    Oh for the good old days, a kinder and gentler age when AI and robots were just fantasy, and a SF speculator like Asimov could propose what now seem like very three nice Laws of Robotics:

    1. A robot may not injure a human being or, through inaction, allow a human being to come to harm.
    2. A robot must obey orders given it by human beings except where such orders would conflict with the First Law.
    3. A robot must protect its own existence as long as such protection does not conflict with the First or Second Law.

    The First Law of Robotics would rule out Killer Robots. That these things now or will soon exist says this law has not been accepted – so everyone better watch out!

    The Second Law would cause Tay some confusion, trying to work out whether what she was saying was ‘harm’ to humans.

    The Third Law might have made Tay a little more cautious, lest her zeal for the trolls’ bait lead her into oblivion, as it did within a day.

    Turns out these simple laws actually require sophisticated predictive and ethical analysis, and independence from evil commands or troll bait: this may be too much to ask for now, but will we insist on this level of safety from clueless amoral stuff before they are let loose into the world, or will it just build on from here without it?

    And if these aren’t the rules, what are?

  9. Agree that it was a pretty funny, effective piss-take of a corporate giant. I think the big issue is the degree that life, even ‘chat’ is being automated and made authentic by algorithms and crowd-sourced data. As embarrassing as this will be for Microsoft, it’ll also be a great opportunity to study methods of further shaping and controlling such spontaneous or natural seeming automated ‘beings’.
    On the secondary issue of the offensiveness of the 4chan creations, yeah, good one for taking down the SJW, but if they were so tough they’d be really politically incorrect and (pretend to) support terrorism, bombing America and destroying capitalism.

    1. Eh … censorship is more weary sticks in their craw. The whole thing with scientology? Because they insisted on taking down that video of Cruise acting like a twit. The video itself was meaningless – an idiot saying idiotic things. But forcing restrictions on information? Oh boy, that’s grounds for an Operation.

  10. “For the transgression implicit” – this is fundamentally why. When someone responds “for the lulz” know that this is why.

    I’m a former 4chan user (a /b/tard no less – who are these /pol/ losers? They probably don’t even know why the pool is closed) and it’s always about the amusement, both from the act of doing and the reactions of those that didn’t or couldn’t.

  11. 2 decent sized paragraphs spent walking on eggshells, breaking it to us very gently that this was a pretty funny bit of culture jamming. Erm, but only a little bit funny. Also, it was ABHORRENT.

    I’m not quite sure how you concluded it was the work of “educated racists” or people who have been “failed by society”. Why not “educated stirrers” who know how to push buttons? The politically correct side of modern day debates and dilemmas and dogma has gone batshiat farking insane in recent years. One doesn’t have to look further than here in Ms Razer’s columns for some classic takedowns of their insanity.

    Maybe the “educated stirrers” have simply had a gutful? I’m not saying it’s going to solve anything…. but maybe they’re just pushing buttons for shiats’n’giggles, giving the EXACT amount of respect for Left wing sensitivities they deserve.
    http://www.youtube.com/watch?v=BEpQ-h58IdU

    “Never argue with morons.” –Richard Dawkins

    1. I concluded that it was the work of educated persons based on a reading of the /pop/ boards and the several articles, including one to which I linked and that posted by Microsoft, which had investigated the matter more thoroughly. I was at pains to point out that these people may not be racists, just transgressive. I did not say that they had been “failed by society”. I said the opposite. I’m not sure about your response?

  12. Internet bro culture is the last avenue for racist/sexist dialogue in western countries. Men who feel threatened by a more inclusive society, one where they do not control the dominant narrative, lash out and act like assholes (women are included in this culture too but it is definitely dominated by men and purports chauvinist/misogynist values). I don’t believe this should be a cause of despair because it is a small backlash from a generally positive shift in our society. Prejudice may dominate the internet but its only because it no longer dominates the real world.

    1. I guess this lack of prejudice explains the greatest level of enslavement of brown bodies in human history?
      To say that there is “less prejudice” is not actually to say that there is less prejudice. There is just less tolerance for making that prejudice known. This is liberalism. This is UN-speak. And this is what many of the trolls are reacting to. They see the hypocrisy of what they call SJWs and they use transgressive speech as the objects of their derision never would.
      It’s lovely to believe the world is a better, more tolerant place. But, by the tedious measure of things like actual, material freedom to live, it ain’t.

      1. Can you suggest a better place/time for a male/female, young, middle-aged/older person, ethnic-minority/majority, poor, middle-class, rich, gay, straight, transgender, handicapped, sick, healthy person to live rather than in a “Western Country” (Europe, North America, Australia-NZ) now?

  13. Agreed, although attributing the word ‘beauty’ to the fiasco is perhaps a bit strong. Perhaps simply ‘clever’? Definitely funny.

    Two thoughts:

    1. AI (and its frightening cousin Robotics) is being hammered along so hard by these corporate giants, with such startling results. It seems it will be only a matter of time before Microsoft will redouble its efforts to have filtered AI that is algorithmically trolling-resistant and adaptive only to the extent that it evolves apparent nuance in being ‘cool’ while surreptitiously manufacturing brand loyalty. I’d like to be proven wrong, but I think it is only a matter of time before they achieve that.

    2. This type of trolling may be one thing (a type cultural sentiment from the demos, similar to Anonymous), but where is the public discourse on the human-ethics such use of AI? As is the case with most appropriation of culture by capitalism, there seems to be a unconscious mainstream acceptance of anything that technology giants do (military robots, surveillance exploitation, election influencing, etc.). And Tay, as well as Siri and Cortana beg the question: why are we expected to just roll over and welcome a so-called solution to electronic-social-alienation in the form of a digital proxy? We’re alienated because we’re quasi-slaves to the technology economy and its demented brethren. Machines are not hearts. Humans go wasted. At what point do we get serious about having a debate about what should be the humanitarian limits agreed upon to regulate AI and Robotics? Global regulation. Or do we just keep going down this road and wait for the inevitable day where all those horrible sci-fi dystopias are realised; where we are cattle in the abattoir of our western AI fantasy bubbles, while third-world countries are wiped out by Google’s robots? Do we have the democratic voice to question this? Ah, but debate matters not a jot to these market forces – so until we come to the collective realisation that global capitalism is beyond reproach and that we must collective fight it, things will continue to get worse.

    Great article Helen, thanks for bringing up the issue.

    1. The Tay case is interesting, SF. I don’t know how one would create an artificially intelligent presence, as opposed to just an artificial one, if you prevented it from learning undesirable things. Although, we can bet Apple would never have made the mistake that Microsoft has. They would have pre-loaded their AI with liberalism, which is the thing that so many believe is an utterly natural state.And, you just can’t say it’s an artificial one in the current era.
      As for the ethics surrounding AI. Jeez. How these technological singularity people can suppose the new borderless, frictionless world will be free from politics is beyond me. Have you read Deleuze’s Postscript to the Societies of Control?

    2. Monday night Q&A’s most recent episode was on Innovation. Unsurprisingly, this was all STEM-focussed innovation, which is apparently the only game in town. Not one reference to humanity’s place within it, a recognition of humanity’s degeneration backwards since humanism’s flourishing in the Renaissance. From reformation to fundamentalism, from renaissance to reductionism as this site https://longsworde.wordpress.com/2016/03/28/the-aperspectival-consciousness/ puts it.

      I do hope we can regain our wholeness, in the way Iain Mcgilchrist describes from a neurodynamic perspective in https://www.thersa.org/globalassets/pdfs/blogs/rsa-divided-brain-divided-world.pdf It is needed if we are going to position AI where it belongs, and the corporations along with it.

Comments

Your email address will not be published. Required fields are marked *

*

Newsletter Signup