This past spring, events promoting Google Gemini, an AI large language model (LLM), popped up on campuses across the US. From Penn to UC Riverside to Howard, the #GeminiOnCampus events included touchscreen kiosk stations demonstrating its capacities for schoolwork, a veritable team of friendly faces in matching tees ready to help you, and, among other freebies, cold-pressed juices as a reward for completing a station.
Having prior worked in the KU Writing Center for 3 years, I have watched generative AI models go from silly novelties that produce nonsensical abstract images to unregulated behemoths capable of photorealistically depicting me committing a crime. I have seen the rise in use of models like Gemini, ChatGPT, and Claude – and the subsequent drop in earnest attempts to learn through writing. I have listened to the people around me talk about how they were pissed because their professor called them out for using AI to write a paper – that is to say, for plagiarizing. As such, I should be clear from the onset: I do not like AI LLMs.
Whenever my dad calls Claude his “friend” or I see a LinkedIn post about the benefits of integrating generative AI into your workflow, I am met with a mix of disdain and sadness. It’s not even really ‘artificially intelligent’; it’s an advanced algorithm. So you can imagine that, when I saw pictures of a giant statue of the Gemini logo outside the KU Ambler Student Recreation Center, my subsequent visit was more of an infiltration than any kind of earnest attempt at engagement.
I drove to the dreaded Lot 90 that morning, hauling several dozen physical copies of The Rose to distribute later. I began mentally preparing myself to suppress my displeasure with AI to see what the event was all about. As I walked up to the staircase, I was temporarily blinded by an unfortunately-angled mirror sign, written on which was a message telling passers-by you could “Find your Spark” with the free juice Gemini was offering. Nearby was a large box truck, telling students Gemini would provide you with “The Finals Boost you Need.”
I was greeted at the door by someone with a bag of squishy Gemini-themed objects. I casually asked him what was going on with the statue, to which he responded by handing me a stress ball, told me that Gemini was here doing an event, and tried to entice me by saying that there was free juice inside. Again with the juice.
Inside, I was met with four large kiosks and a team of around a dozen people stationed around them, all in Gemini shirts. I swiped into the gym and walked over to the line to the ‘event’, which was fortunately rather small. At least three people asked me if I had scanned the QR code, which was plastered everywhere around the area. As it turns out, the code just took me to a Gemini webpage, where it told me to download the Gemini app. I did as I was instructed, though if I had known that doing so wasn’t necessary to receive the full #GeminiOnCampus experience, I would have avoided installing spyware onto my phone
After being funneled into the main area, I went over to the “Focus” station, where I was greeted by two moderately-enthusiastic workers. Despite no one having used the station since I arrived, somehow Gemini – the thing this event was promoting – wasn’t up and running on the kiosk computer, so while one of the workers tried desperately to pull it up, the other made small talk about their theatre degree and how I felt about my history major. Finally, the plagiarism machine was back up and running, though given the demonstration for the station was about how to review for a biology test using flashcards, even if I was interested in using AI, I, a graduating humanities student, still would have found little value in it. Gemini’s ability to generate review materials for me from a Google Doc is something I could do myself with only marginally more effort via Quizlet – or, better yet, with the hundreds of physical notecards I have accumulated through the years. Before departing that station, the workers gave me a few Gemini-themed stickers and – of course – a bottle of vibrant orange juice.
After mustering up enough energy for a second round of this nonsense, I walked over to the “Calm” station, which only had one worker present. This person attempted to explain how Gemini could answer questions about some economics question and could answer follow-ups about its explanation, though neither myself nor the worker could think of anything to ask it. As a pittance for this experience, I received more stickers and – you guessed it – another juice. This one was a dull yellow.
There are plenty of smaller points to raise about this experience. For one, the Rec Center got paid around $10,000 for hosting the event. Combined with labor and material costs, Google paid a comfortable five figures to shill their product to KU students. For another, the event failed to at all clearly or convincingly demonstrate the usefulness or wider applicability of the product on display, both through their ineffectual demonstrations and their bizarre merch choices. Still, there is far more to this discussion than the petty deficits of this chatbot kiosk round robin.
The purpose of this event wasn’t merely to promote Gemini as a particular product; it was part of a broader effort to normalize AI and its growing intrusion into our lives.
No longer can I complete an internet search without AI poorly summarizing the first first few results for me (at least not without a designated Chrome extension). No longer can I scroll social media without seeing soulless AI-generated reaction images, even as I actively fight the uphill battle to purge my timeline of all such content.
Far more importantly, however, no longer can the residents of places from the predominantly-Black city of Memphis to the rural expanse of Utah live their lives without contaminated water, skyrocketing power bills, and the incessant whirring of cooling fans. Moreover, there are serious concerns around the massive data aggregators violating my privacy and breaking me down into a marketized subject, the use of chatbots for companionship and therapy to cope with alienation under capitalism, and the financial instability that results from ‘the economy’ staking its growth on one singular industry. Some are impacted more than others by this technology and all that is necessary to sustain it, but aside from those who own and control it, it is a net negative for all of us.
By setting up shop at campuses nationwide, Gemini is looking to normalize this technology and obscure its costs. Searching with an AI engine, like all other things, is not a neutral process; it does not exist in the aether, detached from the rest of reality – though many mindless consumers have the luxury to behave as though it does. In an age of anti-intellectualism, where critical thought is discouraged and growing frustrations about capitalism are channeled into reactionary causes, AI is the perfect tool by which the ruling class can forestall the development of class consciousness. For many, if AI ‘says’ something is right or wrong, that often provides enough certainty to avoid thinking any further.
But I think there’s more we can learn from #GeminiOnCampus. Like with the event itself, it starts with the juice.
One of the most prominent theories of philosopher and neo-Marxist Slavoj Zizek is that of ideology. From toilets to lawns, our daily lives are fundamentally shaped and permeated by this omnipresent and sublime force. Here, the term refers not to a particular political tradition like conservatism or Posadism; instead, it refers here to the type of thought produced by our surrounding material conditions and systems, which appear as ordinary, mere background noise, or even invisible. While Zizek has written some highly-disagreeable things, he nonetheless offers a solid foundation on which we can understand objects and phenomena through the lens of ideology. Maybe you see where I’m going with this: the juice is ideology.
To press out juice, one must accumulate a great mass and variety of edible matter from around the world, crush it using immense force and energy to extract its concentrated essence, then discard the nutritious ‘waste’ matter, which has, in the eyes of the juicer, served its purpose and is now without use. The #GeminiOnCampus juice was then packaged into disposable plastic bottles, branded on all four sides with Gemini-themed labels, and shipped nationwide to be consumed by some college student and discarded immediately thereafter. In a sense, then, AI LLMs like Gemini are like the juice press; they take petabytes of original, creative, human-made work, extract their essence, discard the source material, and produce a homogenized slop that only resembles the training data in a some qualities, but not in form or content.
The real kicker? The juice tasted terrible! It was supposed to be a mixture of carrot, apple, ginger, turmeric, and lemon, but they got the ratio all wrong and it just tasted like spoiled turmeric, making the resulting liquid worse than the sum of its products. In much the same way, in crushing down every cell of human-made creative works to ‘juice up’ their models in a highly-wasteful manner, AI companies create a product that has none of the tasty benefits of any one of them. By themselves, each of the juice’s components can be good in their own right, but even the most ideal ratio could not possibly retain the crunch of the apple or carrot, the bite of the ginger or lemon, or the depth of the turmeric that make them what they are.
More than a mere analogue, the juice embodies the philosophy of the emerging technocracy in America and beyond. The bourgeoisie’s innate pursuit of profit over all else tends towards suppressing art and free expression which challenges their rule. As such, their control over the process of knowledge production and critical thought can create a barrier between themselves and our endlessly-worsening material conditions. Just as with the juice and AI, the capitalists put the workers through the ringer to extract as much labor as possible, discarding the workers themselves as waste. Thus, even when this extraction makes things worse for most everyone, if it is in the interests of capital to do so, the ruling class will always do it. It would have probably been more expensive to ship fresh produce anyway.
The juice and AI are both means by which we can understand capitalism. Marx writes in Capital about the analysis of a commodity as a way of understanding the system which produced it. AI LLMs like Gemini and their wasteful, harmful, brain-rotting qualities are reflective of the globalized techno-capitalist system that spawned them. As such, to combat AI enslopification, we must combat capitalism itself. To do so, however, we must first understand what we are fighting against as well as how others have fought against similar things. Here, I propose we take inspiration from an unfairly-maligned labor movement from the early days of capitalism in England: the Luddites.
Years ago, I remember learning about the Luddites as a group of misguided, borderline-reactionary factory workers who foolishly destroyed the new machinery in their workplaces meant to make production more efficient. In reality, however, the Luddites were a group of skilled artisans whose labor was displaced by machinery, prompting them to revolt against the capitalists to demand fair treatment. One of the only means by which they could assert their power over the burgeoning British bourgeoisie was through destroying that which threatened to replace them, and so they did. In response, Parliament sent in thousands of troops to repress them, killing many in the process and successfully suppressing their revolt.
I believe we ought to learn from the Luddites and smash the technology that threatens to devalue our labor. If we had democratic control over AI and used it to complete petty tasks while we pursued creative and intellectual passions, we would have no need to engage in any kind of ‘machine breaking’; in a more moral society, the replacement of menial labor would be a great victory. However, because the *alleged* incestual pedophilic rapists and crypto-techno-fascists are hellbent on creating the mythical AI singularity and thereby enshrining a permanent underclass, we must not view AI as anything but a threat to our collective labor power.
AI is a tool, whose moral character, like any other tool, is determined by the hands of its wielder. Targeted use of AI to do things like detecting breast cancer and running protein folding models can yield clearly positive results. Creative use in VFX workflows and voice isolation could enhance rather than stifle the artistic process. Even broader use in self-driving cars and particular chatbots have some promise, should they be deployed only after significant testing and putting the proper safety frameworks in place. There are problems inherent to the technology itself, but many of the issues we see today come not from the nature of AI, but from those who control it and how they are choosing to deploy it.
Still, if it comes down to it, to combat technofascism, we must be ready to launch a Butlerian Jihad against these algorithms of mass destruction. In the Dune series, the Butlerian Jihad was the event in which humans wiped out all thinking machines, taking back the power from the ruling technocratic class in favor of human rule, complete with totally new forms of advanced technology. While the semi-feudal order left in the Jihad’s wake was hardly a liberatory system, we can at least take some inspiration as to what is to be done if we cannot control the technology the ruling class is using today to guarantee their rule for many years to come.
I do not stake my hope for the future on misguided technofixes – and certainly not on the ‘great men’ of our time. Instead, I stake it on a revolutionary spirit that transcends the gilded promises of the technology of the moment and its many investors. If we cannot reclaim the technology sold to us as salvation, we must destroy it and pursue a more human salvation of our own.
Comments
Post a Comment