I’m still getting used to Substack and what it means to publish writing without an editor. Part of the process is learning to see your work from a disinterested perspective, to treat it as a product. This isn’t always easy. It’s sort of like putting yourself into a trance, or maybe breaking a trance.
Recently I’ve been both extremely online and dealing with complicated stuff IRL, which has made it hard to fall into a writing trance. That’s why my last post was so scattered. Mea culpa… and as long as I’m getting meta, I’ll note that I’m experimenting with a new header for subscriber emails (it’s as hand-hewn as it looks!) and that I changed my “about” page to include information about my background, which might be useful for those who don’t know me. Link here.
The essay below is related to a talk I’m giving. I plugged the talk in my last post, but since that’s gone now, I’m going to copy the info to this one. It’s called “The Vanished Flower: How Personal Taste Models the Infinite,” and in part one, I’ll discuss how taste influences philosophical thought, defining “taste” as individual preference for phenomena with sensory properties. As I’ll argue, taste constitutes a kind of knowledge, and this knowledge preserves the sovereignty of philosophy as a domain of intellectual and creative practice, even as LLMs and other genAI tools appear to reveal all forms of thought and creativity to be material and automatable — that is, completely vulnerable to the extractive procedures of information technology.
There’ll be Plato, Kant, Deleuze, and Agamben. Not lightweight stuff, but I’ll use every trick in my teaching bag to make it accessible for the uninitiated. If you have no background in philosophy or AI, it’ll still click. That’s my goal, anyway, while still making it interesting for those who know the literature.
Part two is open-ended. We can talk about art, the sensual, the aesthetic, transformer architectures, whatever we want. I’m hoping participants follow my suggestion to show up with items to share and discuss. Even if they don’t, though, it’ll be fun.
It’s Thursday September 4th, 8-10 pm ET, and it’s hosted by the Weirdosphere learning platform, which is associated with the Weird Studies podcast (a personal favorite of mine). Here’s the full description and registration info, and here’s my Weirdosphere faculty page.
Thanks for letting me advertise once more. On with the essay.
***
I’ll begin with a well-known idea from analytic philosophy, which is that what we call “knowledge” can be defined as “justified true belief.” This definition has been problematized a thousand times over, and it’s never had a lot of purchase in other traditions. I’m starting with it because analytic philosophy is historically entangled with the rise of artificial intelligence, and it’s as useful a point of departure as any.
“Justified true belief” implies that justification, or evidence, isn’t enough to take our thoughts from “belief” status to “knowledge” status. In order to get from A to B, we need to add something called “truth” to the mix. But isn’t that already the point of evidence? Where do we look for truth, if not in the evidentiary data that justify it?
Suppose I guess that it’s raining outside. To confirm or deny this hypothesis, I open the curtains, or I step outside my building. Yes, it’s raining: I see it, I feel it. Isn’t it fair, in this case, to say that I “know” that it’s raining, and that this knowledge encompasses truth? If I believe it to be true, and my senses confirm it, it seems unreasonable to require something further in order to attain “truth” status.
This wouldn’t fly with a lot of philosophers — especially Plato. In Platonic thought, truth lies beyond human awareness and calculation. It’s not that it doesn’t exist, it just doesn’t exist for human beings, at least not in its purest form. Plato recognizes proxies for truth in mathematical forms, and he suggests that we grasp truth indirectly in entities we deem to be beautiful. But there’s no uncomplicated way to invoke it as justification for our knowledge claims. We can’t use truth-in-itself as evidence, since we have no access to it.
The point about recognizing truth in beauty is really important in Platonism. This is because Plato sees attraction to truth vis-à-vis beauty as a necessary precursor to philosophical thought. As Giorgio Agamben explains,
The visibility of the Idea [i.e., truth] in beauty is, in fact, the origins of the amorous mania that the Phaedrus always describes in terms of the gaze and the epistemic process that it brings into being, whose itinerary Plato establishes in the Symposium. There, he characterizes Eros’ stature in the epistemic realm as a medium between wisdom and ignorance and, in this way, compares it to true opinion, knowledge that judges correctly and grasps the truth without, however, being able to justify itself. This medial character of Eros is the basis of its identification with philosophy. (Agamben, Taste, pp. 6-7)
In this passage, Agamben analyzes what Plato calls an “amorous mania” that ensues from the simple and unwitting act of recognizing beauty. This “mania” compels one to attempt to possess the beautiful — to objectify or “freeze” it, so to speak; to make it consistently available to oneself. But before one tries to do this, there’s the moment of perception, a fleeting glimpse that endows one with a very special sort of insight. Plato compares this insight with “true opinion,” or orthe doxa, which is more or less the same thing as an accurate guess.
To describe what Plato’s getting at with “true opinion,” I’ll take a closer look at the case of my unfounded weather prediction. In the first moment, I suppose that it’s raining, even though I have no real reason to believe so — I haven’t yet looked out the window, and we’ll also imagine that I haven’t heard any storm, have no prior knowledge of the forecast, etc. Nonetheless, I think that it’s raining, and as it so happens, I’m right. This belief is a little bit like what one perceives in the moment of recognizing beauty. These are notions that one can’t prove, at least not by recourse to ultimate truth, but they accord with truth nonetheless.
To perceive beauty, Plato tells us, is to love it. And since beauty is a façade for truth, we draw closer to truth as we perceive and subsequently develop feelings of love for the beautiful. Hence love of beauty as philosophy’s proper origin. Etymologically, the philosopher is “the lover of wisdom” — note that in this context, “wisdom” is synonymous with “truth” — so their path always begins with an amorous spark.
This journey, unfortunately, is cursed to interminability. This is because it’s founded on a paradox. As I mentioned earlier, to love beauty/truth is to try to attain it. The problem is that the beauty/truth phenomenon is simultaneously unified and split into two parts — specifically, it comprises two poles connected by what can be thought of as a metaphysical axis, which means that we can only ever move towards one at a time. As we close in on one pole, the methods we use to “freeze” or possess it cast us back towards the other pole, where the exact same thing happens. And on and on.
Truth and beauty are polar opposites because they rule over separate domains. Beauty belongs to the domain of the perceptible, which we might also refer to as the “manifest,” “material,” “embodied,” or “surface.” Wisdom belongs to the domain of the imperceptible, which we might refer to as the “metaphysical” or “ideal.” The latter is the same thing as Plato’s world of forms: it’s the realm of ultimate truth that lies beyond what human beings can directly perceive and cognize.
As an example of the philosopher’s endless journey between beauty and truth, imagine a beautician who wishes to master her craft — to come into full ownership and administrative rule of what strikes her as aesthetically pleasing. Her path might start with the study of cosmetology and proceed to more “pure” forms of science, meaning areas of study that are more theoretical and less applied, such as chemistry. At this point, she’s grasping at pure truths in her attempts to possess beauty as fully as possible.
In the course of gaining scientific knowledge, the beautician notices that the same concepts can be described using different words and mathematical formulas. This is a strange realization, especially as it pertains to the quantitative sciences — the sense that the numbers on a page or screen aren’t identical with the truths they stand for. What the beautician is beginning to apprehend is that one can only access “pure” scientific truths through figures that serve an essentially heuristic function, that is, by approximating something that they themselves are not. With this realization, which is duly philosophical, her quest for truth has brought her back to the world of surfaces, where beauty reigns supreme.
Agamben describes this trajectory as taking place “between a having and a not-having.” The beautician is nothing so much as a medium, an interface between the physical (the beautiful) and the metaphysical (which I’ve described using the words “truth” and “wisdom,” and which constitutes Plato’s world of forms). As she undertakes her voyage, she has philosophical realizations that accord with ultimate truth, such as the notion that quantitative knowledge can be represented with different forms of syntax. As it is, she can’t produce an “ultimate truth” or “pure science” justification for her philosophical ideas: whenever she tries, she changes direction, and the cycle begins again.
—
I want to define taste as the faculty that senses beauty. On this view, taste encompasses what we typically mean when we refer to taste in art or objects with aesthetic qualities, but also taste in its first conventional usage, as in the action that results in the subjective experience of flavor.
Taste is fundamentally preconceptual and prelinguistic. It’s an organ that comes before the thinking mind. It alights on surfaces, allowing one to directly perceive beauty and thus indirectly glimpse the truth. Its role in influencing the rather cerebral task of aesthetic judgment obscures its basic earthiness, its proximity to the sensing body.
Its closeness to the body is what makes taste so particular, so deeply individual. Someone with “good” taste in music or literature can tell you what to listen to or read on the assumption that their experience is representative of some general truth. But I reject the idea that taste expresses anything that applies universally across human experience, and I don’t think that this view deviates from Platonism. (Note that “The Vanished Flower” explores this point in detail).
Because taste is preconceptual and prelinguistic, the various phenomena that appeal to it — the objects and impressions that lead to Plato’s “amorous mania” — aren’t meaningful in a sense that can be directly expressed through language. In the parlance of linguistics, we would say that these phenomena have no innate semantic content. They’re capable of pointing the mind towards meaning that can be rendered as language, but in such cases, there’s no way to know if this meaning was really intrinsic or not. It may have been a clever invention of the thinking mind.
—
I identify the ability to feel love for beauty — to fall into an amorous mania; to taste — as a sort of knowledge in and of itself. It precedes one sort of knowledge and constitutes another. (It’s not “justified true belief,” of course).
One might argue that it’s not fair to frame taste as a form of knowledge since I’ve already framed it as a capacity. To that, I’d say that the supposed difference between capacity and knowledge breaks down under scrutiny.
This scrutiny is more necessary than ever before. Such an examination effectively defends non-scientific bodies of intellectual and creative work — including works in philosophy and everything that counts as art — against claims that there’s no principled difference between their man-made and AI-generated versions.
To tease this argument, AI-generated content is always “about” something. It consists of semantic claims to truth, or information that represents something with which it’s supposedly identical, or which it approximates at least well enough that it’s acceptable to treat the sign and signified as one and the same. Modeled data captures and instantiates a truth that exceeds it, but the whole idea behind data modeling is that the data captures the truth it represents well enough to substitute its real-world counterpart in algorithmic functions (including those that simulate the procedures of “hard” science, like AI-based hypothesis testing).
The content of AI is always just that: content, or forms that we take to represent other forms so well that the difference between representation and referent breaks down. This is a monstrous perversion of the original split between the perceptible and the true (or the manifest and the ideal, beauty and wisdom, etc.). In AI systems, the two domains are collapsed, and the integrity of neither survives. The infinite regress of surfaces leaves us with a void of both; beauty and truth are nowhere to be found here.
Unlike AI-generated content, work in the arts and humanities isn’t necessarily “about” anything. It doesn’t need to contain semantic meaning. It can, of course; much of it does, but there’s a long line of discourse in the theoretical humanities regarding the relationship between representative content and the true meaning of the arts and humanities. Lest I digress…
Freedom from “aboutism” defines the arts and humanities’ proper place. It’s what locates them beyond the areas of research and practice we call STEM. Work that originates in the proper place of the arts and humanities can’t be encoded into other forms, at least not if we want to preserve its metaphysical import. This is one of the major arguments from Walter Benjamin’s “The Work of Art in the Age of Mechanical Reproduction.” Instead of discussing auras, as Benjamin does, I propose a close conjunctive reading of theoretical work on taste, aesthetics, and AI to help us think the difference between art and science anew. That’s what I’m trying to do.
—
I’ve left quite a bit unsaid here. I’m glad that I at least managed to work in my personal definition of “content.” (I’ll defend it vigorously!)
What I’ll be discussing in my Sep 4 presentation is, of course, related to the above, but I haven’t spoiled anything. Plato isn’t the only thinker that matters to this project.
I look forward to seeing you there.