The unconscious is structured like a language.
Jacques Lacan
Over at his Substack blog, American cultural critic and music historian Ted Gioia wrote about his worries that ours may be the last generation capable to distinguish reality from doctored digital fabrications. With the rise of AI and its increased sophistication, Ted warns, very soon it will be impossible to know if videos or photos are genuine or fake.
Examples of AI ‘Deepfakes’
I’m not sure I share Ted’s fears. Sure, my 85-year-old dad is unable to tell whether the photo he received from his friends’ group on WhatsApp is a gen-AI image or not, and he lacks the proper discipline to fact-check the source of the bogus news he finds on his social media feeds; but that’s where I come in—all those years perusing blurry photos of UFOs and Bigfoot finally paid off…
In fact, detecting whether an image is generated by an AI model is arguably getting easier, not harder. Take a look at the more examples of AI slop shared online: have you noticed most of these pictures have a distinct yellow hue in them? This yellowness is not an artistic choice made by the authors (they lack the discernment to make those), and it’s not like AI companies like OpenAI, who developed the most popular AI service (ChatGPT), decided all of the sudden to program their model to incorporate the yellow filter in all the images it churns out, as if they were filming a Breaking Bad episode taking place on the other side of the US-Mexico border.
No, the yellow filter is a telltale sign of a degeneration in the quality of the results provided by the AI services to their users, known in the industry as Model Collapse. It is a vicious and irreversible cycle foretold by critics of these technologies, that occurs when the AI generative models begin to be trained with content generated by other AI models, instead of content created by real humans. As the Internet begins to be littered with more and more AI slop, the newer iterations get more and more degraded until they become useless—like a trapped miner forced to survive off their own bodily waste for lack of other nourishment, they can only do it for so long until it becomes totally toxic.
I still remain highly skeptical of the grandiose promises made by AI techbros announcing that the Singularity is ‘just around the corner’—I’ve been hearing similar predictions made by UFO Disclosure advocates for decades, so my BS detector is more sensible than most—but just like I wrote about the similarities between UFO encounters and the ELIZA effect, I’m still interested in exploring the seemingly peculiar correlations between the paranormal and AI.
Take for instance the early experiments with AI generated imagery, which looked as if the algorithms had been injected with a digital version of DMT: the resulting shapes were fluid, dream-like, chaotic, incongruent, and often morphed human and animal faces—an older gen-AI portrait of Alan Moore made him look like a psychedelic dogman.
Such therianthropic imagery (human figures displaying animal-like characteristics) were very common in ancient petroglyphs, and interpreted by modern scholars as the result of shamanic rituals in which mind-altering substances might have been consumed. But most anthropologists might not be aware that early UFO close encounter reports were also filled with testimonies of witnesses describing similar therianthropic humanoids. Below are a few examples (drawn by yours truly) of such bizarre visitors:
Aside from the animal-headed Ufonauts, there is no denying that the older close encounter cases displayed a wider array of entities reported by witnesses. From leprechauns equipped with gyrocopter chairs—like the one who slaughtered all the sheep of an indigenous woman in Bolivia in 1967—to humanoids with fused legs ending on a single foot, like the ones that invited an anonymous witness to visit their ship in Lomo de Ballena, Perú, in 1949 (the entities told the witness in perfect Spanish they had evolved beyond the need for sex to reproduce, and they could also exchange nutrients just by contact). The ‘mono-foot’ is a trope that would be repeated in other cases, like the famous Pascagoula abduction of 1973, or the less-known abduction from Paciencia, Brazil, in 1977, where the witness was captured and subjected to medical examination by metallic looking robots standing on a cylindrical pedestal, and sporting tentacle-like ‘arms’.
But then came Spielberg’s ‘Close Encounters of the 3rd Kind’, and not long after the intrusion of the so-called alien abduction phenomenon in global consciousness, through the popularization of reports collected and doctored (either through selection or insemination) by Budd Hopkins and David Jacobs.
The cultural impact caused by the cover of Whitley Strieber’s ‘Communion’ has been amply discussed, and cannot be overstated: Before the 1980s, close encounters gathered by investigators from all around the world displayed a colorful bestiary of Ufonauts. After the 1980s, the reports became more amalgamated, until the stereotype of the ‘gray alien’—short stature; emaciated body; oversized, inverted teardrop head; massive black, almond-shaped eyes; slit-like, lipless mouth; barely discernible nose; no ears—became the expected manifestation of alien visitation.
One can try to explain this apparent uniformity with all sorts of speculation. If you’re an advocate of the extraterrestrial hypothesis (ETH) perhaps you would imagine that Earth used to be visited by dozens of alien groups, until the ‘Grays’ drove all the competitors away and imposed a ‘cosmic embargo’ on our planet. We could also imagine that the older, wilder entity configurations were ‘screen memories’ imposed by the ‘Grays’ to confuse the witnesses, but the façade was lifted once researchers learned how to ‘retrieve’ memories from the minds of UFO experiencers—a more parsimonious explanation would be that Hopkins and Jacobs nitpicked the data collected from their experiencer sources, in favor of the narrative they were expecting to find, and discarded the descriptions they deemed ‘confabulations’ because it didn’t fit that narrative.
I like to speculate as much as any other UFO buff—it’s pretty much all most of us can do, due to the lack of actual, verifiable evidence—but since I’m a contrarian by nature and the ETH bores me to tears, I would like to propose another type of speculation as a mental experiment:
Let us suppose that French philosopher Jacques Lacan was right, and the unconscious is structured like a language. Perhaps, by feeding their hungry, hungry algorithms with the entirety of human knowledge available online, AI developers inadvertently created something that is not sentient or even intelligent, as they naively intended, but crudely resembles the landscape of the unconscious mind nonetheless—as Slovenian philosopher Alenka Zupančič proposes in this Substack piece.
But I will go even further than Zupančič, just for the hell of it, and propose that perhaps generative AI models somehow mimic the relationship between the unconscious mind and the paranormal.
As stated previously, we’ve gone from the ‘shamanic’ therianthropes and wild visions of the early close encounters, to the same old, boring ‘flavor of the month’ of the gray alien stereotype—funny how, just like previous AI models often got the number of fingers on human hands wrong, the number of fingers in ‘gray-type’ aliens is also a matter of some debate (some witnesses report three fingers, others four, etc).
If this is indeed the case, and the paranormal is somehow related to how AI algorithms operate, then the current ‘gray filter’ of homogeneous entity reports would be a symptom of that ‘paranormal model’ degrading and collapsing over the decades, after falling into the same vicious cycle experienced by AI models when they are repeatedly trained and re-trained on AI slop. Perhaps the human unconscious suffers the same degradation when it is continuously exposed to derivative UFO slop, churned by generic documentaries and exploitative TV shows.
Do not ask me to provide a mechanism to sustain my hypothesis. As I warned, I am merely engaging in speculation, and I’m light-years away from providing an explanation for the human Id—let alone consciousness itself—functions… just like anyone else.
In any case, I’m more interested in expanding creativity and getting rid of abusive systems of control. I’m not interested in living in a world where OpenAI drives all genuine artistic efforts to the ground, the same way I’m not interested in listening to generic UFO narratives predicated as if they were religious dogmas.
Just as programmers are finding ways to ‘poison’ LLM’s for a variety of reasons—including the protection of artists’ original content—I am wondering now if there is a way to do the same within the paranormal. Using the parlance of my friend Joshua Cutchin’s ‘Fourth Wall Phantoms’, I want to see if there are ways to ‘retcon’ the UFO mythos… although I’m not sure how it could be done.
In a 2023 interview, Computer science visionary Jaron Lanier openly talked about his disagreements over current AI technology with his peers in Silicon Valley. Unlike Sam Altman and most of the techbros, he doesn’t think generative AI models display any sort of intelligence, and warns about ‘worshipping the code’ as if it were the dawn of a coming digital deity.
I feel this is an approach that could prove to be useful in the UFO field as well, where so many ‘worship the alien’ and regard the human component of paranormal experiences as secondary or less important—when in fact it’s probably the reverse.
Just as ‘worshipping the code’ and treating AI models as super intelligences end up dehumanizing society and profiting the promoters of these technologies, ‘worshipping the alien’ has brought us all sorts of unreliable narratives ripe for exploitation by questionable actors. The promises of Disclosure advocates are as ludicrous as those of the Singularity prophets, and they distract us from the real immediate problems we should focus on—climate change and economic inequality to name but a few.
Perhaps it’s time to entertain the disturbing possibility that, just as there is no ghost inside the machine, there might be no higher intelligence behind UFO events or other paranormal manifestations. Perhaps—just perhaps—we’ve been fooling ourselves with the distorted echoes of our own unconscious voice.
