The Gimmick
Draft: Still working on this piece, bunch of messy notes halfway through. With longer pieces like these, my thoughts are constantly changing and I want to make the thinking in public part apparent.
I came across the writings of Sianne Ngai while looking for texts and reading material on cuteness, for another post I’m working on that began with me parsing through the dissonant feelings of why I mildly dislike my work/self being called cute (like when someone says, oh how cute), while at the same time, really liking and finding comfort in cute things, and investigating where these feelings might emerge from, something I suspected was in relation to ideas around gender, performance, agency, and existence under capitalism…
This post is a tangent from that post.
While going through Ngai’s work, I came across the idea of the gimmick, the central theme of her third book Theory of the Gimmick: Aesthetic Judgment and Capitalist Form (2020) and was immediately struck by the parallels I could draw to the current discourse on the widespread adoption of artificial intelligence (more specifically, LLMs) in personal and professional work, despite the book not mentioning A.I. once.
Sianne Ngai is a literary cultural theorist who writes about aesthetics under current capitalist culture, and how the ambivalent feelings and judgments certain aesthetic objects elicit can tell us about the cultural conditions of our world, and how we process its hyper-commodified, information-saturated nature.
She describes the gimmick as:
“The gimmick is an object – any commodity, not necessarily an artwork – that strikes us as aesthetically suspicious because of the way in which it seems to be working either too hard or too little. Our experience of its impoverished form triggers evaluations of excessive or deficient labour, which highlight the object’s false claims to value.”
She argues that the gimmick appears to save labour, while simultaneously generating more work - and that our suspicion and ambivalent feelings towards these contradictory objects and their claim to value directly speak to our views on labour, time, and value under capitalism. She is also directly interested in our aesthetic judgment of these objects, which is instantaneously formed almost in conjunction with the affect (the emotional response generated in us by something external) that the form of these objects have on us.
“The gimmick is thus capitalism’s most successful aesthetic category but also its biggest embarrassment and structural problem. With its dubious yet attractive promises about the saving of time, the reduction of labor, and the expansion of value, it gives us tantalizing glimpses of a world in which social life will no longer be organized by labor, while indexing one that continuously regenerates the conditions keeping labor’s social necessity in place.”
I find this to be eerily on the pulse about our current misgivings (and on the other side, givings) with artificial intelligence. The hype, the backlash, the mild aversion, the curiosity, the defeated acceptance, the critical nay-sayers, the early adopters… all of us seem to be grappling with its far-reaching implications on an individual level, which speak directly to our ideas on time, labour, and value. How does one make sense of something that shakes up your perception on how things should work? How does this bring into light the larger cultural, infrastructural, and global forces that surround us?
…In our everyday encounter with the gimmick, we are thus registering an uncertainty about labor—its deficiency or excess—that is also an uncertainty about value and time.
The judgments of labor, time, and value encoded in the gimmick suggest that this aesthetic category reflects nothing less than the basic laws of capitalist production and its abstractions as these saturate everyday life.”
I have not read any of Ngai’s work in its entirety, and have largely used this interview with the author, and the introduction of this book as a reference. My context is someone from an urban Indian context who does creative/design work, makes art, and consumes popular online tech and culture discourse. This largely speaks to white collar professions. I’m interested in seeing how some of the common themes that come up in this discourse map to Ngai’s idea of the gimmick.
Working very little, but also very hard
LLMs give us immediate ease: fast answers, immediate payoff, condensing whole days of work into minutes. This goes against so much of how we think of work.
At the same time, there is so much hidden work happening to generate this apparent ease: electricity and water and other physical resources used for data centers, the human labour that goes into the data AI is trained on, the clickworkers who contribute to training data, all the resources used for hardware and more.
Here, the gimmick is that the AI can do things with extreme speed and ease, seemingly replacing whole jobs and industries, with what looks like no effort, potentially replacing human labour… while also being simultaneously driven by human labour and work, as well as consuming huge amounts of physical resources to train data and function, enroaching lands, contributing to unethical web scraping and copyright infringement.
Another side to this: Does it actually reduce the human work? Or instead, is more work going into managing and making sure these systems work correctly?
In the context of software development, this is a common complaint one can see across forums, especially developer forums who are constantly dealing with vibe coded apps and end up having write things from scratch anyway, not sustainable. Or even, complaints about the quality of their own colleagues and the code they write.
It is true that this is an era of figuring these systems out: questions of developing workflows that can sustain themselves and work with AI, etc. It is still work, with the promise of it being the future, and you just need to do the work now, since this is going to be it. Get with it, or get left behind.
This also brings me to a common framing of a certain sect of wholehearted AI proponents. It’s not AI, it’s just you who’s thinking of it wrong, participating in it wrong, not optimizing your usage. Your intrinsic worth is tied to your output in this system, and if you are not using it properly, then you are reducing your own value in a professional marketplace.
Understandably, this brings up mixed feelings in many of us.
But even if AI is the future and all of this is true, and
Then what?
Another common theme that comes up is that the point of AI is to automate away the annoying bits so that you can do work that you care about, that actually matters. Menial, grunt work comes up in everyone’s profession. It is usually the job of the juniors in a traditional workplace/profession to take this up, till they can progress to higher roles where they can do more of the thinking, managing, and high level work.
looking at it through lens of consumption, existing within a society where it almost makes sense to use ai, gen z approach to work.
I think the gimmick plays out in a few flavours here.
Does automating the annoying bits away actually give you time to work on the things you care about? Do you end up doing that? Or do you end up spending time managing the system, or simply doing absolutely nothing since you can now?
Under capitalism, there is no escape: automate away the annoying work so that you can do more work, until that work becomes the work that can be done cheaper with AI, and then… where does it end? It all seems to be generating more and more work, especially if we are looking at work under the current modes of production. wary of the gimmick but at the same time seeing how neatly it works and fits and contributes to the context it resides in.
A third view - in avoiding the ‘grunt’ work, what are we losing out on?
Before, we were told to see value in grunt work, see it as necessary to learn the foundations. Of course, one wants to get out of this position as fast as possible. But now, this has been decimated as we are told that grunt work simply needs to be done by AI - simply because it can be done cheaper.
is it a bad thing for ‘grunt work’ to be automated away leaving you to focus on what you want to do. is it a bad thing to use ai for quick money making while you can actually work on what you want to do?
ai is even helpful across workflows, from design to coding to data analysis. but knowledge of the unimportant work is what gives one a certain command and confidence with their tools. or does it?
new ways of knowing, working that arise from this is natural. learning to drive manual vs automatic, the implicit knowledge that manual learning first is better than automatic. not a rite or a ritual thing, but when handling a machine gives you confidence, allows you to do more?
(Also, to note, a lot of this ‘menial work’ tends to be the jobs that have been given to women in a lot of industries.)
Value with respect to human labour
Nowhere does our ambivalent reaction come out more than it does with AI generated art and its aesthetics.
We question the intrinsic value of AI generated art, design, code, products - since there was no ‘work’ at all, is it really valuable? How does one define value in the first place? Aversion to AI generated art and visuals, especially when off training data of actual humans, is the value in the content created reduced because a human didn’t spend days, weeks, making it?
Also just visually off.
Maybe to you or me, it matters that a human made something, that there is value because it is human. But this is a bit taut, since there are things I have considered of little value that have been made by humans.
I think the framing is more in respect to the fact that now that AI does exist, and that businesses will go for whatever is seemingly cheaper, to extract as much value as they can: aesthetics, products, commodity, the gimmick, art clashing with market and capitalism as opposed to art in its own sphere - hence aesthetics, capitalising off them. It was never about how the output was made in the first place, just that it got made, unless /how it got made/ was a selling point.
at the same time, we question the value of ai art and imagery since there was no ‘effort’ - says something about how we give value to things based on the labour that goes into them, though art tends to generally be the first to be devalued. human labour intrinsically ‘worth’ more now, or is it? art in the marketplace context by itself is inherently not doing much in a capitalist society except for being valuable etc, so human labour might be something that actually gives it more value to continue to give more value in a capitalist society. <<–– not sure
^^ from an aesthetics POV: generating visuals using specific training data to get specific aesthetics etc, commodification of these aesthetics further from the reasons they were originally created and contexts they were created in, abstracted, is there no value because it was generated from a bunch of training data and it is kind of pretty? how are people consuming this?
Human thought actually worth more now, in a design context: taste, but being able to drive an AI.
Good ideas can sustain any trend that comes.
….Genuine value
I am not trying to frame this as an anti-AI post. I am genuinely trying to parse through the opinions I encounter in discourse, with people around me, as well as the internet, and look at the two sides of the whole thing.
I have found value in using LLMs to assist me with coding projects I could have never undertaken without a lot of friction, work, time, and understanding, as well as assist me with tasks that would have just taken longer to do other wise. I don’t know if this saves time, since I am not fully taken with the optimising my life with AI side of things, neither am I trying to jump on the /I build things with AI/ wagon (yet). I simply find it useful for learning and making and looking for information. I am curious about the possibilities of things I would never have seriously thought about otherwise.
At the same time, fully aware and have noticed the effects of regularly conversing with LLMs, cognitive offloading, less and less of a feeling of command and authority and understanding over my work.
making non-technical people, technical and vice versa: also a gimmick? emptiness - lack of understanding code, or creativity: playing at something one does not possess, presented as ones own work nevertheless. emptiness.
AI has enabled people who were anyway tinkering with code, say, to do things they would never have forayed into. I do think there is genuine genuine value in enabling people. There are all the infrastructural caveats, but I am do not want to discount personal benefits.
Though I think in this case, the gimmick is, what am I losing in this process that doesn’t seem to be doing much harm, and instead much good?
Ease!= value
Output!= understanding.
I am not the first to actually look at Ngai’s work in relation to artificial intelligence, unsurprisingly. I came across this hackernews discussion thread and this article that argues against LLMs being of any real value in higher education.
more and more research articles about how ai usage messes with our ability to critically think, overestimate our abilities, talks to us like a sycophant. how this changes how we engage, think
recognizing the gimmick, being aware of the claims both ethical and moral and societally while also for your own personal engagement with the world. what else do you have than how you live, love, learn, show up, contribute, exist?
i believe that you can only really start with the personal and the community around you. infrastructual forces, markets, economies the way these work –> of course affect everything. i also dont think everyone can keep up with this information, having a place for discourse and conversation is important.
awareness community finding and figuring out your own personal goals in the context of the larger
for me:
i love designing and building visual, moving, interactive narratives. i love the internet, growing up, stumbling across things organically, everything it has shown me and given me that contributes to how i view the world. i love working with people to make things that would not have existed before. i absolutely love the process of figuring something out, moving continents to make things happen, and the adrenaline click when it does.
most important now to have communities to see through the gimmicks together, open spaces for discussion around ai usage. to use, to be critical, to learn how to talk to machines, to understand what exactly it is giving you, to have spaces for asking questions and learning, instead of being alone in your room talking away to sycophantic LLM, sticking with hard bits.