Willy's Simulacrum Experience
What the Glasgow Wonka fiasco and Baudrillard tell us about AI's relationship to the real
“We used artificial intelligence to generate those [images]”
“Is it a fake?”
“No, absolutely not.”
Last weekend, an event billed as “Willy’s Chocolate Experience” in Glasgow used AI generated images to hype a magical (if trippy and glitchy) world of animals and candies and sweetened topographical features—although admitted that “This experience is in no way related to the Wonka franchise.” What they delivered—for 45£—was a sparsely decorated warehouse walk-through with tacked-up backgrounds and painted stairs, paid actors hired days before and given AI-generated scripts, and, for each child: one quarter cup of lemonade and one jellybean. Parents were pissed.
As long as you’re not one of the poor parents or kids who experienced “Willy’s Chocolate Experience” (which apparently included zero chocolate), there’s a lot to laugh at here. The juxtaposition between the promises and reality here is amazing meme fodder.
But of course I’m here for the AI part. The images on the website for Willy’s Chocolate Experience (click now before they take the site down!) are obviously AI-generated copyright infringements of the most recent Willy Wonka movie.
If you look too closely at the Enchanted Garden, you will see some tortured and frightened bunnies. Maybe a clue pointing to the actual Glasgow warehouse experience?? One commenter looking at the actual photos of the event noted, “I would 100% think my organs were being harvested if I showed up to that place LMAO.”
I admit that as a parent, I would be most excited about the Twilight Tunnel:
In the 'Twilight Tunnel,' , get ready for an exhilarating and immersive adventure. Journey through a dimly lit passage adorned with captivating projections, enigmatic sounds, and surprising turns that will immerse you in suspense and excitement. It's a heart-pounding experienceyou've never experienced before!
It's the UKXEPCTED TWITS that I would really relish.
Although who doesn’t love a “cartchy tun” or a “pasadise of sweet teats”? Paging Lewis Carroll! Or what one Twitter/X user disturbingly called a “bearussy.”
OK, enough with the easy dunks on AI. The point is that the event didn’t and couldn’t come close to what the images promised.
What’s actually interesting to me here is an assertion the event organizer makes in this video. Go ahead and watch it, but good luck parsing it if you’re not Scottish. Here’s the relevant transcription, around 20sec into the video:
“Is that why you don’t have pictures of all your stuff?”
“No”
“Actual, photoshops of your stuff?”
“No. We used artificial intelligence to generate those [images]”
“Is it a fake?”
“No, absolutely not.”
“Is it a fake?”
(Six months in Glasgow in the late 90s came in handy for me here, as well as a Scottish friend—thanks, Mairi!)
The event organizer claims that the event(?)/the images(?) are not fake because they used artificial intelligence. They didn’t use Photoshop to alter the images from some actual, grounded reality. Instead, with AI, they created something wholly disconnected from reality. It wasn’t a fake.
It was a simulacrum.
So, of course, I had to go back and read Baudrillard. It’s been a minute.
And the section of his Simulacra and Simulations that I found most relevant is about “The divine irreference of images:”
…feigning or dissimulating leaves the reality principle intact: the difference is always clear, it is only masked; whereas simulation threatens the difference between ‘true’ and ‘false’, between ‘real’ and ‘imaginary’.
Baudrillard uses the example of someone feigning symptoms of a medical condition. If you’re feigning ill, then you’re just pretending, and presumably could be found out with a proper medical examination. But if you’re simulating illness, you’re actually producing some of the symptoms of the illness, and it’s not so simple whether you have it or not. That’s why simulation is much more dangerous to our concept of reality.
And of course, simulation is what AI does. The AI images used to advertise Willy’s Chocolate Experience have no reference in reality—except that they’re some weird amalgamation of Warner Bros intellectual property and a diffusion model of image generation. They aren’t fakes. A Photoshop, or fake, would leave the reality principle intact.
Baudrillard goes on to describe the “successive phases of the image:”
It is the reflection of a basic reality.
It masks and perverts a basic reality.
It masks the absence of a basic reality.
It bears no relation to any reality whatever: it is its own pure simulacrum.
With AI images, we’re in stage 4: pure simulacrum.
What’s even better here is that AI drove the event itself: the actors’ scripts were AI-generated impossibilities: “Aside from featuring staging and effects that would be impressive for a moderately-sized West End spectacular, its stage directions also dictate precisely how delighted the audience will be.”
The actor playing “Willy” rehearses some of this garbage in a video where he laments that it’s all now in his brain.
Another actor played the AI-generated “evil chocolate maker” The Unknown and scared children at the event—although I think she redeemed herself in this video in which she deadpans: “Children were crying; there’s no denying that. But if you could see into their eyes like I could, you would have seen that they were understanding this as a powerful piece of theater.”
Hilariously, the event organizer is also the “author” of AI-generated scammy anti-vax conspiracy books, which have since been pulled from Amazon. Perhaps this proves the point that AI isn’t the problem: it just supplies more tools for scammers to use. Still, I think the utter disconnect with reality that these AI-generated images implies (and perhaps the vax conspiracy books as well) suggests a new phase of hyper/un/reality that will take some getting used to.
As the event organizer tells the customer accusing him scamming parents and kids, if you’ve read this far, I’ll say:
“On behalf of the House of Illuminati…Thank you for your time.”
P.S. Just for fun, here is some Baudrillard juxtaposed with images from Willy’s Chocolate Experience. Since I hit the file limit for Substack, I posted more on Twitter.
I love your pairings of Baudrillard quotes with the images. In a previous moment (like three years ago?), we would have a Twitter bot generating Baudrillard/Wonka scam image pairings, but that was killed by two things: Musk and so-called "generative AI." Musk's destruction of the Twitter API for bot makers was a harbinger of things to come, I think. The real nail in the coffin of the fun days of Twitter bots is GPT. Thinking about people delighting in Twitter bots is almost like watching a Comedy Central standup special from 1994. You just wonder how anyone could have been laughing at it? The old school, bespoke Twitter bot makes little sense in a world where GPT exists.
This is fabulous, Annette. My favorite of course is "...AI isn’t the problem: it just supplies more tools for scammers to use. Still, I think the utter disconnect with reality that these AI-generated images implies (and perhaps the vax conspiracy books as well) suggests a new phase of hyper/un/reality that will take some getting used to." 100%. And we get a whole lot of entertainment in the process.