"I don't write about AI"--I find myself saying this to friends and colleagues with some frequency. And yet the topic is difficult to avoid these days. Below are some notes on one of the fundamental problems in AI: Is is possible for a discrete machine to model continuous phenomena? Instead of trying to solve the problem--indeed I suspect it is unsolvable--computer scientists might simply stipulate it as an axiom. Every discipline rests on a series of fiat decisions. Why not AI as well?

One of the first things I teach in my undergraduate course on "Creative Coding" is the concept of data type, and, more specifically, the distinction between two primitive types, integer and float.

The integers are, of course, the arithmetical counting numbers; they are whole values and advance by discrete steps: 4, 5, 6, 7... Whereas floats (short for floating-point numbers) represent values with a decimal or fractional component, for example 0.1 or 3.141592.

Informally I instruct my students to think of integers as "digital" numbers and floats as "analog" numbers. Of course that's not technically true, since both integers and floats are discrete, finite values; and indeed both are themselves represented by integers (confirming the truism that all digital computers are, in fact, digital). But it's true in spirit. Integers are chunky and monadic, like digital units. While floats are quasi-continuous and fine-grained, like analog magnitudes. Continue reading →

During my explorations into the digital and analog, I often encounter a series of frequently asked questions. I will address three of them here...

Question: does it have to be either/or? Must one decide between the digital and the analog? What about digital-analog mixtures? What about ambiguity?

No, it is not necessary to decide. Both/and is a fine solution. There exist endless hybrids and alloys of the digital and the analog. For instance, the analog may be rigorously characterized as a "twin" or "echo" of the digital, rather than its rival or contrary.

However, also note that ambiguity and mixing are characteristic of analogicity. Favoring such things is the analog position par excellence. Hence it is a mistake to assume that the blur or the mixture or the hybrid will escape the digital-analog duality, when, in fact, such endorsements merely reinforce one of the two terms within the duality (the analog). In order to withdraw more definitively from the digital-analog system we will need much more powerful techniques than blurring, mixing, or hybridizing. Continue reading →

Several common mistakes should be avoided when trying to define the digital.

(1) The first mistake is to assume that the digital is bound by the field of consumer electronics.

Focusing on consumer electronics tends to confine digitality within a relatively narrow band of history (industrial societies roughly since World War II) and a relatively small group of media objects (computers and their ilk).

(2) Another mistake is to assume that the digital means binary.

While the binary is an excellent digital technology, and binary distinction is the first digital event, any number of other digital technologies operate effectively with more than two binary tokens. The Latin Alphabet uses 26 digital tokens rather than two. The decimal integers use 10 digital tokens rather than two. A grand piano uses 88 digital tokens rather than two. In other words, the binary is not an adequate explanation for the digital; the binary itself must be explained.

(3) Likewise it is also incorrect to assume that the digital necessarily means computation.

Computation involves a mechanism that processes values or signals, given finite resources. We ought to differentiate between a type of representation and a type of calculation. The digital is a mode of mediation; it is not a mechanism, at least not immediately. And of course analog computers exist, along side digital computers. The question of the digital can never been fully explained by the answer "computer."

(4) A final error to be avoided is to assume that the digital means data or information. Indeed information may be represented digitally or analogically, depending on the case.

That said, the mere existence of a piece of information (whether digital, analog, or another mode) entails a provisional digitality, this being the just-noticed difference that demarcates a piece of information as different from its context. (Such provisional digitality--the bare fact of difference as yet unstandardized--is what I formally call "Digital 1." The event of  converting provisional digitality into a regular discrete framework is called "Digital 2.")

Thus when speaking about data or information, it is important to consider two aspects, first the evidence of any form of difference whatsoever (a.k.a. just-noticed difference, or Digital 1), and second the specific mode of representation being deployed (continuous variation, discrete units, or perhaps some other mode entirely).

In the past I've written about the importance of cuts within digitality. A cut indicates difference, difference of whatever kind. But a cut is not enough for a full-fledged digital system. Cuts will not spontaneously assemble into a regular series, and cuts will not calculate or compute, at least not if they remain in the larval stage of heterogenous difference. An additional operation is necessary to convert a cut into the basis monad for constructing a regular discrete framework. A number of names have been proposed for this second event: semioticians describe the emergence of a master signifier, Marxists refer to the selection of a general equivalent, in contemporary finance the name is tokenization. In each case, something valuable is converted into a new form where it may be expressed through, and measured in terms of, a specific representational basis monad.

Jean-Joseph Goux penned a useful summary of this operation in his important 1968 essay titled "Numismatics," first published as two parts in late 1968 and early 1969, where the refrain "Gold, Father, Phallus, Monarch, Language" announced the scope of his ambition.

By the mid 1960s Althusser had warned readers of Marx to avoid the opening sections of Capital, vol. 1 due to their purported philosophical excesses, a Hegelian admixture being the chief hazard. Yet Goux brazenly rejected Althusser's advice--and we shall follow his example--knowing that Marx's theory of value, which opens Capital, was absolutely crucial. Continue reading →

Join me for an event at NYU with Martha Schwendener about her new book The Society of the Screen: Vilém Flusser’s Radical Prescience. We will talk about Flusser, art and art making, screens, technical images, cybernetics, and many other things.

April 9, 6:30pm
Einstein Auditorium
34 Stuyvesant Street
NYC
Register

Digital theory will benefit from a return to structuralism. As I've said before in print, structuralism was the apex of digital theory during the 20th C. During that period, scholars were obsessed with codes, logics, binarisms, linguistic structures, and even quasi-mathematical approaches, as seen in the work of Claude Lévi-Strauss, or Lacan's Seminar 2 on cybernetics and binary codes. Scholars like Lydia Liu and Bernard Geoghegan have traced some of these lineages; Leif Weatherby's new book, Language Machines, takes up the question of structuralism as it relates to AI. I value their work and we ought to learn from them.

Meanwhile, problems linger, problems that attend any "return" to structuralism. I will pass over all the well-known shortcomings, already chronicled in various ways over the years: the risks of over-investing in the synchronic (at the expense of the diachronic); the rigidity of an apparatus frozen in amber (as opposed to a more dynamic or internally contradictory model); the tendency for structure to reify binary relations; structuralism's apparent Eurocentrism. Those problems remain. I'd like to address some additional shortcomings, though, ones that spring directly from digital theory. Continue reading →

While the four of us largely agree on the basic terms discussed in Digital Theory, a few specific points of disagreement remain. Dialectics is one of those points of disagreement. Handelman and Weatherby's chapter, titled "Digital Dialectics," argues that the two terms are intertwined. Meanwhile, my chapter insists that the digital is strictly incompatible with dialectics. The argument is simple: the digital militantly observes both the principle of identity and the principle of noncontradiction, age-old logical principles known since Aristotle at least; whereas the dialectic constantly violates those two principles (which is why we love it). Continue reading →

I'm excited to announced that Digital Theory is now published in both book form and open access PDF. I haven't actually seen a physical copy yet.. but it's apparently out in the world.

The four of us--Beatrice Fazi, Matthew Handelman, Leif Weatherby, and me--recently went on both the Disintegrator and Acid Horizon podcasts to discuss what it might mean to think the digital theoretically and theory digitally.

Here are a few highlights:

-- looking beyond the "consumer electronics" theory of the digital

-- the return to structuralism

-- Deleuze as analog philosopher. (I've discussed this before.)

-- disagreement over the status of the dialectic. We didn't have time to discuss this on the podcast, but it's a minor theme in the book. I'll try to post more thoughts about this in the future.

So tired of "gen" in AI/computation/etc.

Hoping for new horizons of degen in 2026.. full-stack permadeath.. crypto-erasing.. stunlock design patterns.. fractal death drive..

full-stack permadeath -- one and done. you babies.

crypto-erasing -- asset forfeiture, guaranteed. like when the K Foundation burned a million quid.

stunlock design patterns -- for (int i = 1; i > 0; i++) { /*the fun begins here*/ }

fractal death drive -- the compulsion to repeat, at all levels of scale.

inverse Nyquist -- always set sample rates to less than half of audible frequencies.

actually, fuck it, 1-bit compression for everything -- that email you sent this morning? on/off. the King James Bible? on/off. all of Wikipedia? on/off.

SAaaS (Solar Anus as a service) -- convert the cloud to run on The Amnesic Incognito Live System.

hyper-GEN -- greygrey googoo.

one-time-pad alphabets -- `sudo kill -9 -1` but for Ferdinand de Saussure.

let's get ycombinator going on some of these ideas!!

Recently I've been revisiting the cinema of cuckoldry. You know, those films where a roving Lothario colludes with a married woman to cuck-n-kill her husband before getting whacked himself. There's Double Indemnity (1944), of course, and The Postman Always Rings Twice (both 1946 and 1981 versions). There's the peerless Body Heat (1981), where the woman wins in the end. There's Framed (1947), albeit with the essential coordinates slightly tweaked. Something about the mid '40s and the early '80s synched well with the logic of the cuck.

Forty years on, we're well into another cuck wave. I'm not thinking cinema so much as subjectivity and perhaps even society as a whole. Who are today's cucks exactly? I will refrain from the typical alt-right potshots: soy libs are cucks; wife guys are cucks; watching mainstream media is cucked; et cetera. (Spelling out etc. is extremely cucked, btw.)

Still, it helps to be specific. Those people who go jogging while making hands-free phone calls with a Bluetooth earpiece, those people are cucks. The cucks are the 100k follower Twitter accounts. The cucks are the Opinion writers in the Washington Post. The cucks are the military personnel who voted for Trump (three times). But, interestingly, Trump is also a first-class cuck, perhaps for different reasons. A proper cuck would be Kanye saying flattering things about Adolf Hitler. More cucked is Elon striking a Nazi pose several months later. And even more cucked are those downstream from Elon, his followers and enablers. The cybercucks bought Cybertrucks. We could go on and on through this pitiful roster. In fact, did you know that Charles Fourier once wrote a treatise defining 72 different types of cuckold? Continue reading →