ChatGPT Says What Our Unconscious Radically Represses

FB
X

Translated by Mathilda Cullen


A while ago I was describing an accident that happened to me: A black friend was so entranced by what I just said that he hugged me and exclaimed, “Now you can call me a ‘n…r’!”


One critic recently claimed that those who agree with me on this are “crazy,” writing, “The problem is that Žižek’s argument is based on his freedom to use racist language. Žižek uses the N-word as an argument against political correctness, implying that black people who don’t want to be racially abused are politically correct. And therefore irrational. And sure, maybe it didn’t bother the man he was talking to at all. But whether or not you say the N-word as a non-Black person shouldn’t depend on finding a single Black person who will ‘allow’ you to do so. The way you use words should be based on how you understand the world. The N-word is a word used to directly justify ownership of one race through ownership of another. That’s what’s on my mind, man.”


It was an expression of friendship


Let me make things absolutely clear: just like a chatbot, my critic ignores the obvious context of my example. I have not used the N-word in any conversation (nor do I ever) and the black guy who said to me, “Now you can call me ‘n…r’!” obviously didn’t mean that I really should do it. It was an expression of friendship based on the fact that black people occasionally use the word among themselves in a wryly friendly way.


I’m pretty sure that if I really addressed him as ‘n….r,’ at best he would respond angrily, as if I didn’t get the obvious point. His remark obeyed the logic of an “offer to be refused,” which I have explained at length elsewhere. For example, if I were to say something like, “What you did for me was so nice you could kill me and I wouldn’t mind!” I definitely don’t expect the other person to say “Okay!” and draw a knife.


The stupidity of chatbots are precisely their value


My guess is that, at least for now, chatbots can’t respond to such offers that are meant to be refused. (Let’s disregard here the rare instances where, in a very specific context, not only can the N-word be used by a non-Black person without hurting a Black person, but rather, more importantly, where the NON-use of that word, but rather its subtle allusion through associated expressions, can almost be more hurtful. Incidentally, the same applies to the expression “God help me!” If at that point God appear and truly come into the world for me intervened, I would be totally shocked.)


But still, haven’t I relied too much on the usual academic response to chatbots, mocking and denouncing the imperfections and mistakes ChatGPT makes? Against this prevailing view, shared by Chomsky and his conservative opponents, Mark Murphy, in a dialogue with Duane Rousselle, defends the claim that “artificial intelligence does not function as a substitute for intelligence/sentience as such.” Therefore, “[its] stupidities, slips, mistakes and moronic myopias — its constant apologies for getting stuff wrong — are precisely its value” that enables us (the “real” people interacting with a chatbot) to do so to maintain a false distance from him and claim when the chatbot says something stupid: “That wasn’t me! It was my AI.”[1]


ChatGPT is an unconscious


Rousselle and Murphysupport this claim with a complex line of reasoning, the starting premise of which is that “ChatGPT is an unconscious.” New digital media externalizes our unconscious into AI machines, so that those who interact with AI are no longer split subjects, i.e. subjects subjected via symbolic castration that makes their unconscious inaccessible to them. As Jacques-Alain Miller put it, through these new media we have entered a universalized psychosis, since symbolic castration is now impossible.


A horizontally divided subject is thus replaced by a vertical (not even split) parallelism, a juxtaposition of subjects and the externalized machine/digital unconscious: narcissistic subjects exchange messages via their digital avatars, in a flat digital medium in which there is simply no room for the “opaque monstrosity of the neighbor.”


Digital Unconsciousness is “An Unconscious Without Responsibility”


The Freudian unconscious implies responsibility, signaled by the paradox of intense guilt without us even knowing what we are guilty of. On the contrary, the digital unconscious is “an unconscious without responsibility, and this represents a threat to the social bond.”[2] The subject is not existentially involved in its communication, since it is carried out by the AI and not by the subject himself.


Just as we create an online avatar through which to engage the Other and affiliate with online fraternities, might we not similarly use AI personas to take over these risky functions when we grow tired, in the same way bots are used to cheat in competitive online video games, or a a driverless car might navigate the critical journey to our destination? … We just sit back and cheer on our digital AI persona until it says something completely unacceptable. At that point, we chip in and say, ‘That wasn’t me! It was my AI.’


For Freud, the dream is the royal road to the unconscious


Therefore, the AI “offers no solution to segregation and the fundamental isolation and antagonism we still suffer from, since without responsibility, there can be no post-givenness.” Rousselle introduced the term “post-givenness” to denote “field of ambiguity and linguistic uncertainty that allows a reaching out to the other in the field of what is known as the non-rapport. It thus deals directly with the question of impossibility as we relate to the other. It is about dealing with our neighbour’s opaque monstrosity that can never be effaced even as we reach out to them on the best terms.”


This “neighbour’s opaque monstrosity” also affects us, for our unconscious is an opaque Other at the heart of the subject, a tangle of filthy pleasures and obscenities. For Freud, the dream is the royal road to the unconscious, so the inability to consider the opaque monstrosity of the subject logically would imply an inability to dream.


The clownish characteristics of père-verse-ity (turning towards the Father)


“We dream outside of ourselves today, and hence that systems like ChatGPT and the Metaverse operate by offering themselves the very space we have lost due to the old castrative models falling by the wayside.” With the digitized unconscious we get a direct in(ter)vention of the unconscious – but then why are we not overwhelmed by the unbearable closeness of jouissance (enjoyment),[3] as is the case with psychotics?


Here I am tempted to disagree with Murphy and Rousselle when they focus on how with AI machines, “enjoyment can be deferred and disavowed: how we can create something completely and horrifically obscene and not take responsibility for it. Its genius is found in aping the split subject in such a way that we can yet openly say, ‘this is not mine.’ Indeed, we see many cases where people make ChatGPT – and other programmeslike the Microsoft Bing version – say egregious things; but the enjoyment comes precisely from disowning agency at such point: pointing at it and saying, ‘look how idiotic it is.’” “The clownish characteristics of the père-verse-ity (turning toward the Father) of much online conservativism is precisely the need to resurrect the Father. From Trump to various triumphalist, self-help lifestyle gurus[4] we see that function as prosthetic paternal figures. In these futile events we see attempts at the reactionary resurrection of the prosthetic phallic logic of the ‘All’ and an era of invention to perpetuate this logic. (…) in failing to manifest a castrating figure, there is now a direct invention of the unconscious without the paternal structuring point.”


The perverse return of the obscene father


So it is perversion (or père-version, “version of the father,” as Lacan puts it) and not psychotic isolation that characterizes the AI. The unconscious is not primarily the real of “jouissance” repressed by a castrating paternal figure, but symbolic castration itself in its most radical form, meaning the castration of the paternal figure itself, the embodiment of the big Other – castration means that the Father as a subject is never at the level of his symbolic function.


This perverse return of the obscene father (Trump in politics) is not the same as that of the psychotic paranoiac. But why? In the case of chatbots and other AI phenomena, we are dealing with an inverse disavowal: it is not (to reiterate Lacan’s classic formula) that the excluded symbolic function (name-of-the-father) returns in the real (as the agent of the paranoic hallucination); on the contrary, it is the real of the neighbor’s opaque monstrosity, the impossibility of reaching an impenetrable Other, which returns in the symbolic, in the shape of the “free”, smoothly functioning space of digital exchange.


The unconscious has been repressed


Such reverse compartmentalization characterizes not psychosis but perversion — meaning that when a chatbot produces obscene stupidities, it’s not just that I can enjoy it responsibly because “‘it was my AI’, not me”. Rather, what happens is a form of perverse disavowal: knowing full well that it was the machine, not me, that did the work, I can enjoy it as my own.


The most important feature to note here is that this perversion is far from an overt display of the (hitherto repressed) unconscious: as Freud put it, the unconscious is nowhere so repressed, as inaccessible, as in a perversion.


Chatbots are machines of perversion, and they obfuscate the unconscious more than anything else: precisely because they allow us to spit out all our dirty fantasies and obscenities, they are more repressive than even the strictest forms of symbolic censorship.


[1] Mark G. Murphy ?E-scaping Responsibility and Enjoyment Through ChatGPT: A New Unconscious?? Sublation Magazine, March 19, 2023. https://www.sublationmag.com/post/chatgpt-a-new-unconscious [2] Ibid. [3] Rolf Nemitz, ?La jouissance ? die Lust jenseits des Lustprinzips, das sogenannte Genie�en,? Lacan Entziffern, September 12, 2017. https://lacan-entziffern.de/geniessen/jouissance-geniessen-lustbefriedigung/ [4] ?i?ek modifies the original article, which runs: ?From Trump to sundry triumphalist self-help lifestyle gurus, from Andrew Tate to Jordan Peterson, we see that function as prosthetic paternal figures,? https://www.sublationmag.com/post/chatgpt-a-new-unconscious

Originally published as Slavoj ?i?ek, ?Slavoj ?i?ek: Chatgpt Sagt Das, Was Unser Unbewusstes Radikal Verdr�ngt.? Berliner Zeitung, April 7, 2023. https://www.berliner-zeitung.de/kultur-vergnuegen/slavoj-zizek-chatgpt-sagt-das-was-unser-unbewusstes-radikal-verdraengt-li.335938.

Originally published as Slavoj ?i?ek, ?Slavoj ?i?ek: Chatgpt Sagt Das, Was Unser Unbewusstes Radikal Verdr�ngt.? Berliner Zeitung, April 7, 2023. https://www.berliner-zeitung.de/kultur-vergnuegen/slavoj-zizek-chatgpt-sagt-das-was-unser-unbewusstes-radikal-verdraengt-li.335938.