Generative “Artificial Intelligence” & Psychoanalytic Writing: An Editorial Note
If we begin with the premise that at the very least, nearly all recent psychoanalytic publications, including those in this journal, have availed themselves of generative “artificial intelligence” (any type of AI system capable of generating text, images, or other media in response to user prompts), then we will have said both nothing new and will have provoked a worthwhile and necessary rupture with certain fictions we hold dear.
With few deliberate exceptions, we are ALL now caught in the rapture of the machine. To wit, some small portion of this text was composed nearly-automatically by my word processer, feeding me ideas on how to end my sentences (predictive text), based on my previous compositions, that of others, and the programming of its engineers. Again, nothing new. This AI creep is something we are dimly aware of, even if we don’t want to think about it too much; in fact, generative AI, even at this relatively simple level, is explicitly designed to help us think less, about everything, including our writing.
However, the ante has been radically raised with the advent of more recent large, language model chatbots, like ChatGPT, capable of generating extensive, systematic, and internally coherent texts. As submissions arrive at the offices of EJP, how are we as editors to determine the provenance of any given text? Again, some portion of it was most certainly machine-written, just like this note. But what if the entire text was machine-generated? Or substantial pieces of it, if only to add flourish and finish? Surely, we could tell the difference! As it turns out, we sometimes cannot. Just try it for yourself. Sign up with ChatGPT (or any of the other myriad number of chatbots scurrying into view) and ask it to compose an essay on the political works of Alfred Koestler in the style of Gabriela Mistral. The results go beyond mimicry. Judge for yourself.
Now, I should hasten to add that I certainly do not think that generative “AI” is indeed smart or “intelligent” in the way humans are. And I categorically disagree with the delusional notion that generative AI is “conscious” or even “self-aware”. And yet, “writing” produced by these machines born of enormous computing power, access to nearly infinite data sets, and simultaneous and instantaneous learning across bots, are original texts, of a kind. How are we to treat them? What do they portend for speech and language? And how do they enter (if at all) into the domains of psychoanalytic writing and symbolic exchange? I will grant that these questions seem somewhat preposterous. But we must ask them, or others will answer them for us.
Now, our answers are not immediately forthcoming. It might be some time before we can even begin to provide a response adequate to the concerns at hand. It could be that we never even get to do so. But again, it is vital to formulate our questions now, to stem the machinic onslaught even for a bit, so as to find some kind of footing in the midst of these digital torrents.
Towards that end, I do want to mark the advent of this latest iteration of generative AI as a watershed moment. We have now gotten to the point where we (including the generative AI programmers themselves) simply do not entirely understand what these bots are doing, what they might be capable of executing in the near term (especially if they are more intimately paired with other systems, such as financial markets, war-machines and utility infrastructures) and what the impact on human thought and practices will be. If we bear in mind the unabashed avarice of the Silicon Valley outfits leading the generative AI charge, the future looks ghastly indeed.
Some might regard my thoughts on this as alarmist. At this point, the only retort I could give to that contention would be to ask: what else would it take to sound the siren? It is now the case that our capacity to establish or ascertain veracity is impaired on at least two fronts. One: we are increasingly less capable of determining the factual truth of anything that is written or reported. Two: it is becoming progressively more difficult to determine the derivation of any supposed statement of fact. To put it together: it is getting damn difficult to determine who or what wrote or said a thing, and whether that some-thing is factually accurate or not. Without these two faculties, where are we? Some might like to float in the indeterminacy of the realms we are hyper-catapulting into. But psychoanalysis, as a field, demands something of a different order. This is not to say that “facts” are what analyses are made of. Quite the contrary. But the field, via its manifold forms of literature and modes of transmission does need to ascertain veracity in different and complex ways. This is now in serious jeopardy.
With that in mind, we invite our readers to submit texts that speak to the questions raised above. We are hoping to publish more articles and essays on this urgent set of problems, and we might even dedicate a special issue or Salon section to the effort. Just make sure that it is you, however you understand that to be, that is writing the text. Or at least, make us believe that it was human-generated. It takes the unconscious to lie. To my knowledge, there is no way of programming that into a chatbot as of yet. Generative AI may blur the line between a lie and the always fraught effort at truth-speaking, but at least for now, lying is solely the reserve of humans.
In this new issue, we begin our psychoanalytic forays with a captivating essay that speaks directly to some of the social and subjective transformations being wrought by these newer articulations of artificial intelligence. In The Stainless Gaze of Artificial Intelligence: A Lacanian Examination of Surveillance and Smart Architecture, author Marc Heimann and contributor Anne-Friederike Hübener consider how artificial intelligence, under the guise of “smart cities” and “smart homes”, may very well be usurping the place of the big Other. Playing with Joan Copjec’s formulation of “gaze as stain”, Heimann and Hübener’s inversion of the same is used to argue for the advent of a new kind of panopticon that reduces modern subjectivity down to its calculable properties.
Shifting gears, we turn to Judith Beyer’s delightful Perversion and the State: Lacan, de Sade, and Why 120 Days of Sodom is Now French National Heritage. Encouraging us to take up Lacan’s teachings on perversion yet again, Beyer’s incisive theorizing argues that by virtue of purchasing Sade’s book and declaring it a part of the national heritage whilst simultaneously making it inaccessible, the French state has not only attempted to rid the text of its transgressive character (thereby making of it a fetish object), it has also made a bid to prop itself up while struggling to mask a void. The Name-of-the-Father as well as the (mis)fortunes of patriarchy also receive a treatment by Beyer that is worthy of study and further elaboration.
The only summary that could ever be given of a poem is yet another one. As such, I offer you the one provided by the author of our third piece himself: “Since birth was given…debt(h). What knowledge is taken as. Flight.” Pronounced by none other than our own Stephen Mosblech, his contribution to this issue, Hate: An Obsessional Phantasmagoria features La Gioconda and Leonardo in a rapturous, intoxicating brew. Drink up my friends! This is too human a piece to have been given birth by a chatbot.
We round out the articles in this issue with a profoundly important and worthwhile treatise from a rising voice, Ashwak Hauter. In the course of taking up many different arguments and drawing from both psychoanalysis and direct anthropological fieldwork, her piece, Reconstructing the Community, Reconstructing the Image: Refuge in Islam in Yemen and Lacan After Islam manages to take an effective and sophisticated position contra many previous attempts at theorizing the connection between psychoanalysis and religion. Eschewing the notion that religious traditions and practices can be simply reduced to the registers of sublimation or identity politics, she argues (among other things) for the enormous value of Islam’s manifold traditions in helping both patient and analyst alike to steer a course through the riven sociopolitical and economic landscapes they inhabit.
As a sort of enlivening interlude, the issue continues with an absorbing interview of the philosopher Daniel Tutt by Anthony Ballas. Zeroing in on how fundamental concepts in psychoanalysis, particularly the Oedipus complex and the theory of the superego, necessitate reformulation in light of tectonic shifts at the level of the family, the conversation proceeds to cover all manner of terrain that is of use for both the clinic and social thought.
This issue features five book review essays; a development in line with EJP editorial efforts to survey and comment on the prolific output of analytic texts in English. We kick off this section of the issue with a lengthy and detailed review of Todd McGowan’s The Racist Fantasy: Unconscious Roots of Hatred by Division/Review editor-in-chief, Loren Dent. McGowan’s worthy tome comes in for both praise and some critique for the way in which the relationship between enjoyment, fantasy and identification is understood.
We continue with our own editorial board member Nicolle Zapien’s appreciative and incisive review of Mary Brady’s edited book, Braving the Erotic Field in the Psychoanalytic Treatment of Children and Adolescents. Featuring several heavy-weights in the world of field-theory, the book advances approaches that speak directly to how work with the erotic in the analytic treatment of young people can take place.
Frequent EJP contributor and fellow editor, Christopher Chamberlin, chimes in with a luminous and lucid review of Karima Lazuli’s Colonial Trauma: A Study of the Psychic and Political Consequences of Colonial Oppression in Algeria. Clearly impacted by his reading of Lazuli’s work, Chamberlin boldly uses the opportunity of the book review to ask challenging questions regarding post-colonial and “non-colonial” subjects in psychoanalysis that we would be wise to attend to.
Philosopher and EJP editor Ed Pluth brings his many talents to bear on Alenka Zupančič’s Let Them Rot: Antigone’s Parallax. An erudite and sophisticated review, Pluth points out the many gifts of the text, working with Zupančič’s arguments so as advance a number of inspired articulations regarding a whole host of issues, including the deadlocks and quandaries of contemporary subjectivity.
We round out the issue with long-time EJP editor Matthew Oyer’s refined and penetrating appraisal of Bret Fimiani’s Psychosis and Extreme States: An Ethic for Treatment. Reading with and eventually through Fimiani’s gentle style, Oyer lays bare the teeth of the book and underscores the importance of the work for the field at large.
Before ending, a quick note about upcoming items. We release our second issue of 2023 at the end of the year, which will also see the publication of our newest Salon Feature, “Against the Grain: Emergent Psychoanalytic Perspectives on Autism”. Early 2024 brings yet another Salon Feature, “Climate Catastrophe and Psychoanalytic Practice”. We welcome inquiries and submissions for any of these two new Salon Features as well as our upcoming Winter issue.
By way of concluding, I would like to thank our esteemed editorial board, our generous and dedicated peer reviewers, Managing Editor Stephen Mosblech and each of the contributors to this issue. As well, I extend my gratitude to our two interns, Phyllis Metzdorf, who handled the lion’s share of copy-editing for this issue and Esteban Jijon, who undertook the copy editing for our recent Spanish issue and continues to apply his labors to the Salon side of the journal.