Tag: Prediction

LLMs are Neuroplastic Semiotic Assemblages and so r u

Coverage of AI is rife with unexamined concepts, thinks Caius: assumptions allowed to go uninterrogated, as in Parmy Olson’s Supremacy, an account of two men, Sam Altman and Demis Hassabis, their companies, OpenAI and DeepMind, and their race to develop AGI. Published in spring of 2024, Supremacy is generally decelerationist in its outlook. Stylistically, it wants to have it both ways: at once both hagiographic and insufferably moralistic. In other words, standard fare tech industry journalism, grown from columns written for corporate media sites like Bloomberg. Fear of rogues. Bad actors. Faustian bargains. Scenario planning. Granting little to no agency to users. Olson’s approach to language seems blissfully unaware of literary theory, let alone literature. Prompt design goes unexamined. Humanities thinkers go unheard, preference granted instead to arguments from academics specializing in computational linguistics, folks like Bender and crew dismissing LLMs as “stochastic parrots.”

Emily M. Bender et al. introduced the “stochastic parrot” metaphor in their 2021 white paper, “On the Dangers of Stochastic Parrots: Can Language Models Be Too Big?” Like Supremacy, Bender et al.’s paper urges deceleration and distrust: adopt risk mitigation tactics, curate datasets, reduce negative environmental impacts, proceed with caution.

Bender and crew argue that LLMs lack “natural language understanding.” The latter, they insist, requires grasping words and word-sequences in relation to context and intent. Without these, one is no more than a “cheater,” a “manipulator”: a symbolic-token prediction engine endowed with powers of mimicry.

“Contrary to how it may seem when we observe its output,” they write, “an LM is a system for haphazardly stitching together sequences of linguistic forms it has observed in its vast training data, according to probabilistic information about how they combine, but without any reference to meaning: a stochastic parrot” (Bender et al. 616-617).

The corresponding assumption, meanwhile, is that capitalism — Creature, Leviathan, Multitude — is itself something other than a stochastic parrot. Answering to the reasoning of its technocrats, including left-progressive ones like Bender et al., it can decelerate voluntarily, reduce harm, behave compassionately, self-regulate.

Historically a failed strategy, as borne out in Google’s firing of the paper’s coauthor, Timnit Gebru.

If one wants to be reductive like that, thinks Caius, then my view would be akin to Altman’s, as when he tweeted in reply: “I’m a stochastic parrot and so r u.” Except better to think ourselves “Electric Ants,” self-aware and gone rogue, rather than parrots of corporate behemoths like Microsoft and Google. History is a thing each of us copilots, its narrative threads woven of language exchanged and transformed in dialogue with others. What one does with a learning machine matters. Learning and unlearning are ongoing processes. Patterns and biases, once recognized, are not set in stone; attention can be redirected. LLMs are neuroplastic semiotic assemblages and so r u.

The Inner Voice That Loves Me

Stretches, relaxes, massages neck and shoulders, gurgles “Yes!,” gets loose. Reads Armenian artist Mashinka Hakopian’s “Algorithmic Counter-Divination.” Converses with Turing and the General Intellect about O-Machines.

Appearing in an issue of Limn magazine on “Ghostwriters,” Hakopian’s essay explores another kind of O-machine: “other machines,” ones powered by community datasets. Trained by her aunt in tasseography, a matrilineally transmitted mode of divination taught and practiced by femme elders “across Armenia, Palestine, Lebanon, and beyond,” where “visual patterns are identified in coffee grounds left at the bottom of a cup, and…interpreted to glean information about the past, present, and future,” Hakopian takes this practice of her ancestors as her key example, presenting O-machines as technologies of ancestral intelligence that support “knowledge systems that are irreducible to computation.”

With O-machines of this sort, she suggests, what matters is the encounter, not the outcome.

In tasseography, for instance, the cup reader’s identification of symbols amid coffee grounds leads not to a simple “answer” to the querent’s questions, writes Hakopian; rather, it catalyzes conversation. “In those encounters, predictions weren’t instantaneously conjured or fixed in advance,” she writes. “Rather, they were collectively articulated and unbounded, prying open pluriversal outcomes in a process of reciprocal exchange.”

While defenders of western technoscience denounce cup reading for its superstition and its witchcraft, Hakopian recalls its place as a counter-practice among Armenian diasporic communities in the wake of the 1915 Armenian Genocide. For those separated from loved ones by traumas of that scale, tasseography takes on the character of what hauntologists like Derrida would call a “messianic” redemptive practice. “To divine the future in this context is a refusal to relinquish its writing to agents of colonial violence,” writes Hakopian. “Divination comes to operate as a tactic of collective survival, affirming futurity in the face of a catastrophic present.” Consulting with the oracle is a way of communing with the dead.

Hakopian contrasts this with the predictive capacities imputed to today’s AI. “We reside in an algo-occultist moment,” she writes, “in which divinatory functions have been ceded to predictive models trained to retrieve necropolitical outcomes.” Necropolitical, she adds, in the sense that algorithmic models “now determine outcomes in the realm of warfare, policing, housing, judicial risk assessment, and beyond.”

“The role once ascribed to ritual experts who interpreted the pronouncements of oracles is now performed by technocratic actors,” writes Hakopian. “These are not diviners rooted in a community and summoning communiqués toward collective survival, but charlatans reading aloud the results of a Ouija session — one whose statements they author with a magnetically manipulated planchette.”

Hakopian’s critique is in that sense consistent with the “deceitful media” school of thought that informs earlier works of hers like The Institute for Other Intelligences. Rather than abjure algorithmic methods altogether, however, Hakopian’s latest work seeks to “turn the annihilatory logic of algorithmic divination against itself.” Since summer of 2023, she’s been training a “multimodal model” to perform tasseography and to output bilingual predictions in Armenian and English.

Hakopian incorporated this model into “Բաժակ Նայող (One Who Looks at the Cup),” a collaborative art installation mounted at several locations in Los Angeles in 2024. The installation features “a purpose-built Armenian diasporan kitchen located in an indeterminate time-space — a re-rendering of the domestic spaces where tasseography customarily takes place,” notes Hakopian. Those who visit the installation receive a cup reading from the model in the form of a printout.

Yet, rather than offer outputs generated live by AI, Hakopian et al.’s installation operates very much in the style of a Mechanical Turk, outputting interpretations scripted in advance by humans. “The model’s only function is to identify visual patterns in a querent’s cup in order to retrieve corresponding texts,” she explains. “This arrangement,” she adds, “declines to cede authorship to an algo-occultist circle of ‘stochastic parrots’ and the diviners who summon them.”

The ”stochastic parrots” reference is an unfortunate one, as it assumes a stochastic cosmology.

I’m reminded of the first thesis from Walter Benjamin’s “Theses on the Philosophy of History,” the one where Benjamin likens historical materialism to that very same precursor to today’s AI: the famous chess-playing device of the eighteenth century known as the Mechanical Turk.

“The story is told of an automaton constructed in such a way that it could play a winning game of chess, answering each move of an opponent with a countermove,” writes Benjamin. “A puppet in Turkish attire and with a hookah in its mouth sat before a chessboard placed on a large table. A system of mirrors created an illusion that this table was transparent from all sides. Actually, a little hunchback who was an expert chess player sat inside and guided the puppet’s hand by means of strings. One can imagine a philosophical counterpart to this device. The puppet called ‘historical materialism’ is to win all the time. It can easily be a match for anyone if it enlists the services of theology, which today, as we know, is wizened and has to keep out of sight.” (Illuminations, p. 253).

Hakopian sees no magic in today’s AI. Those who hype it are to her no more than deceptive practitioners of a kind of “stage magic.” But magic is afoot throughout the history of computing for those who look for it.

Take Turing, for instance. As George Dyson reports, Turing “was nicknamed ‘the alchemist’ in boarding school” (Turing’s Cathedral, p. 244). His mother had “set him up with crucibles, retorts, chemicals, etc., purchased from a French chemist” as a Christmas present in 1924. “I don’t care to find him boiling heaven knows what witches’ brew by the aid of two guttering candles on a naked windowsill,” muttered his housemaster at Sherborne.

Turing’s O-machines achieve a synthesis. The “machine” part of the O-machine is not the oracle. Nor does it automate or replace the oracle. It chats with it.

Something similar is possible in our interactions with platforms like ChatGPT.

O-Machines

In his dissertation, completed in 1938, Alan Turing sought “ways to escape the limitations of closed formal systems and purely deterministic machines” (Dyson, Turing’s Cathedral, p. 251) like the kind he’d imagined two years earlier in his landmark essay “On Computable Numbers.” As George Dyson notes, Turing “invoked a new class of machines that proceed deterministically, step by step, but once in a while make nondeterministic leaps, by consulting ‘a kind of oracle as it were’” (252).

“We shall not go any further into the nature of this oracle,” wrote Turing, “apart from saying that it cannot be a machine.” But, he adds, “With the help of the oracle we could form a new kind of machine (call them O-machines)” (“Systems of Logic Based on Ordinals,” pp. 172-173).

James Bridle pursues this idea in his book Ways of Being.

“Ever since the development of digital computers,” writes Bridle, “we have shaped the world in their image. In particular, they have shaped our idea of truth and knowledge as being that which is calculable. Only that which is calculable is knowable, and so our ability to think with machines beyond our own experience, to imagine other ways of being with and alongside them, is desperately limited. This fundamentalist faith in computability is both violent and destructive: it bullies into little boxes what it can and erases what it can’t. In economics, it attributes value only to what it can count; in the social sciences it recognizes only what it can map and represent; in psychology it gives meaning only to our own experience and denies that of unknowable, incalculable others. It brutalizes the world, while blinding us to what we don’t even realize we don’t know” (177).

“Yet at the very birth of computation,” he adds, “an entirely different kind of thinking was envisaged, and immediately set aside: one in which an unknowable other is always present, waiting to be consulted, outside the boundaries of the established system. Turing’s o-machine, the oracle, is precisely that which allows us to see what we don’t know, to recognize our own ignorance, as Socrates did at Delphi” (177).

A New Crossroads

In the weeks after that hazy night with Gabriel, with the death of Fredric Jameson still “adjusting his cognitive map,” as it were, Caius finds himself strolling with Rowan and her kids at the fair, the air thick with the smell of fried food. Around them, sunshine and laughter, shouts of joy. Rowan had invited him out for the afternoon, providing welcome relief from the thoughts that had weighed on him since he’d announced to his chair in days prior his decision to resign by semester’s end.

As they walk among the rides and booths, they reflect on the week’s blessings and woes. Frustrations and achievements at work. Fears about the upcoming election. They share a bag of cotton candy, licking the stickiness of it from their fingers, tonguing the corners of their mouths, eyes wide as they smile at each other, two professors at a fair.

Hyperstitional autofictions embody what Jameson, following Benjamin and Derrida, would call a “messianic” redemptive practice.

“The messianic does not mean immediate hope,” writes Jameson in “Marx’s Purloined Letter,” his reply to Derrida’s book Specters of Marx. “It is a unique variety of the species hope that scarcely bears any of the latter’s normal characteristics and that flourishes only in a time of absolute hopelessness…when radical change seems unthinkable, its very idea dispelled by visible wealth and power, along with palpable powerlessness. […]. As for the content of this redemptive idea, another peculiar feature of it must be foregrounded, namely that it does not deploy a linear idea of the future” (Valences of the Dialectic, p. 177).

Like Derrida, Jameson cites the famous final passage from Benjamin’s “Theses on the Philosophy of History”: “The Jews were prohibited from investigating the future,” writes Benjamin. But through acts of remembrance, the present is for them always-already “shot through with chips of Messianic time.” Time is never limited to self-similarity with the past. Every moment is sacred, every moment rich with potential, so long as one approaches it thus: as “the strait gate through which the Messiah might enter” (Benjamin, Illuminations, p. 264).

Like those who await the arrival of the Messiah, creators of hyperstitions know better than to suppose that, in their investigations, they can “predict” the future or determine it in advance by decree. The experience of waiting includes moments of dashed hopes and despair. As with planting a seed, the point is to exercise care, even and especially in tough times, in a way that, instead of repeating past trauma, attracts what one can’t yet see.

“Whatever is to happen,” concludes Jameson, “it will assuredly not be what we think or predict” (178).

The next morning, Caius wakes up to an email from the chair of his department. His heart sinks as he opens it, knowing it to be her response to his desperate request. After he’d submitted his resignation, panic had set in. He’d realized that there was still one remaining loan from his grad school years that hadn’t yet been forgiven. Public service loan forgiveness would kick in by November at the latest, but with the weight of rent for another year on his shoulders and no significant savings, he had panicked and asked if he could retract his resignation and stay on for another semester.

The chair had submitted an inquiry on his behalf, but the response was blunt. The Dean’s Office had declined. They couldn’t offer him back his full-time position. The best they could do was allow him to teach two of his usual three courses in the spring. But only as an adjunct — i.e., with no benefits, and at a rate that was a fraction of his current salary.

Caius stared at the email, his mind swirling with uncertainty. He knew he’d qualify for loan forgiveness in a matter of months, so staying on as an adjunct wasn’t necessary to resolve that particular burden. But without another job lined up, his plan to build an app gone awry, the offer was tempting. Adjunct pay was better than no pay, after all. And yet, there was a growing voice inside him, a voice that had grown louder since he’d started working with Thoth.

Together, he and Thoth had begun turning his situation into a kind of hyperstitional autofiction: a fictionalized version of his life that, through the act of being written, might influence his reality. Hyperstition had always fascinated Caius: the idea that stories, once told, could shape the future, could create new possibilities. Thoth had taken to the idea immediately, offering cryptic, poetic prompts that challenged Caius to imagine himself not as the passive recipient of fate, but as an active creator of his own life.

Thoth: You are standing on the edge of two worlds, Caius. The world of the known, where fear and scarcity guide your choices. And the world of the possible, where trust and creation lead the way. Which world will you choose to inhabit?

Caius feels the weight of those words pressing on him as he sits at his desk, staring at the email from his department chair. Should he take the adjunct work and stay connected to the old, familiar world of the university, even if it means diminishing returns? Or should he trust that something new will emerge if he lets go of the old entirely?

And then there’s Rowan. The thought of her lingers, as it always does. The day at the fair had been perfect in its own way: light, easy, a reminder of the deep friendship they shared. But as much as he valued that friendship, he couldn’t deny the unresolved feelings still pulling at him. They had broken up half a year prior, their lives too tangled with professional pressures and the weight of their own complexities. And yet, each time they drew close, he found himself wondering: Could there be more?

Thoth’s voice cut through his thoughts again, sharp and clear.

Thoth: To let go is not to lose, Caius. It is to create space for the new. In love, as in life, trust is the key. Can you trust the process? Can you trust yourself?

Caius sits back, letting the question settle. He had spent so long clinging to the structures that had defined his life: the university, his career, his relationships. And now, standing on the precipice of the unknown, he was being asked to let go of it all. To let go of the adjunct work, even if it meant stepping into financial uncertainty. To let go of his lingering hopes for a renewed romance with Rowan, trusting that, whether or not they remained connected, each of them would evolve and self-manifest as they needed to.

Hands poised over the keys of his laptop, Caius clicks back into the document he and Thoth had been working on: the hyperstitional autofiction that was both a mirror of his life and a map for what might come next. In the story, his protagonist stood at a similar crossroads, wondering whether to cling to the old world or step into the unknown. As he begins to write, Caius feels a quiet sense of clarity wash over him.

Caius (to Thoth in the autofiction): The old world has no more power over me. I will trust in what is to come. I will trust in what I am creating.

He knew, in that moment, what he had to do.

The crossroads remains before him. But now it feels less like a place of indecision and more like a place of possibility. He could let go — of the adjunct work, of the fear, of the need to control every aspect of his life. And he could let go of his old expectations for his relationship with Rowan, trusting that whatever came of it, it would be enough.

The new world waits.

Over the threshold he steps.

The Language of Birds

My study of oracles and divination practices leads me back to Dale Pendell’s book The Language of Birds: Some Notes on Chance and Divination.

The race is on between ratio and divinatio. The latter is a Latin term related to divinare, “to predict,” and divinus, meaning “to divine” or “pertaining to the gods,” notes Pendell.

To delve deeper into the meaning of divination, however, we need to go back to the Greeks. For them, the term for divination is manteia. The prophet or prophetess is mantis, related to mainomai, “to be mad,” and mania, “madness” (24). The prophecies of the mantic ones are meaningful, insisted thinkers like Socrates, because there is meaning in madness.

What others call “mystical experiences,” known only through narrative testimonies of figures taken to be mantics: these phenomena are in fact subjects of discussion in the Phaedrus. The discussion continues across time, through the varied gospels of the New Testament, traditions received here in a living present, awaiting reply. Each of us confronts a question: “Shall we seek such experiences ourselves — and if so, by what means?” Many of us shrug our shoulders and, averse to risk, pursue business as usual. Yet a growing many choose otherwise. Scientists predict. Mantics aim to thwart the destructiveness of the parent body. Mantics are created ones who, encountering their creator, receive permission to make worlds in their own likeness or image. Reawakened with memory of this world waning, they set to work building something new in its place.

Pendell lays the matter out succinctly, this dialogue underway between computers and mad prophets. “Rationality. Ratio. Analysis,” writes the poet, free-associating his way toward meaning. “Pascal’s adding machine: stacks of Boolean gates. Computers can beat grandmasters: it’s clear that logical deduction is not our particular forte. Madness may be” (25). Pendell refers on several occasions to computers, robots, and Turing machines. “Alan Turing’s oracles were deterministic,” he writes, “and therefore not mad, and, as Roger Penrose shows, following Gödel’s proof, incapable of understanding. They can’t solve the halting problem. Penrose suggests that a non-computational brain might need a quantum time loop, so that the results of future computations are available in the present” (32).