Edit: I updated this significantly in August of 2025 to include some new information and examples that occurred since this was originally written in May 2024.
I lurk a bit on Reddit's Pagan and Witch forums, and today I saw something... disconcerting.
Somebody posted a couple of images from a set of artwork they had purchased for scrapbooking. Not wanting to use something with text they didn't understand, they posted to a couple Witch-related pages to ask for a translation. These are the pictures they posted:
You probably already know what the problem is, it's in the title after all: There is nothing to translate. The text is gibberish that was spat out of an AI image generator. People quickly pointed this out, but were quick to add "...but you know, it's still cute though."
Since I originally wrote this—thank Gods—I've noticed a sharp turn against AI in the Pagan community, with most of the forums I frequent banning its intentional use (obviously the above person did not know it was AI, she was the victom of a scam). But there are still Pagans and Witches out there running to its defense, or using it in their projects or (even worse) things they sell to other Pagans.
AI is a problem. OK, I'm not saying every use of AI is bad—I'm not going to tell you that you can't use it to put together a shopping list or break down your chores into smaller chores—but generative AI used to create "art" and "literature," or even to provide factual information, has created serious issues, not only for artists and writers who are having their work plagiarized and their livelihood threatened, but for people looking for accurate, credible information. And although it's a widespread problem that affects everyone, I am particularly concerned about how this affects the Pagan and Witch community, for a variety of reasons I'm going to talk about today.
I am very eclectic in my practice. I understand that my magical ancestors—the ones who practiced folk magic, who stuck with "superstitious" traditions, the weird Catholics lighting candles and burying statues in the front yard to try selling their houses, all the way back to the ones painting plump animals with tallow and blood in caves—did not stick with one specific cultural practice. They adapted to changing technology by innovating new ways of using magic. Many Witches—including myself—continue to do it this way, and many of us start doing things a certain way and then completely forget why we ever started doing it that way to begin with.
This can cause problems. I started practicing Witchcraft as a literal child based on information I found on the Internet in the late 1990s, and a lot of the things I did at this time became codified in my brain as just "how you do things" that later turned out to be problematic, poorly-researched, or just nonsense. As a 12 year old with no prior experience with Witchcraft, some of the things I learned to do were literally just sarcastic jokes Witches wrote at the expense of other Witches that I didn't get because of my lack of experience. So I was this 12 year old who had a binder full of information printed off websites, including "Ancient Egyptian Love Spell: Place a photograph of him under your pillow" and a "How to create a craft name" guide that was just a joke about Silver Ravenwolf and people who followed her work. A while ago somebody developed a fake Kabbalah circle meaning "tomato soup" to make a point about cultural appropriation and using symbols you don't understand, which a lot of people did in fact take seriously.
This sort of thing is not limited to 12 year olds' Trapper Keepers and Kabbalah appropriators, either. It is not uncommon to find things in Pagan, Wiccan, and Witchcraft books that are completely uncited and don't make any sense. If you've ever read the commonplace piece of advice that you can substitute rosemary for any other herb in any spell, I'm like 99% certain this proliferated because Scott Cunningham wrote it in a book, where it was published as a fact with no context, citation, or reasoning. Somebody then put that on the Internet, and it became copied and copied and copied to the point where a lot of people just take it as common knowledge. To this day I've been trying to track down where this advice originally came from, and so far I have found nothing before that Scott Cunningham book. It's just a completely context-less piece of advice with no rhyme or reason at all that everybody just hears at some point and takes at face value.
But here's the thing: Despite my own personal problems with these incidents, all of these are actual practices people do. It is common (and in fact not really a big deal) for people to generate "Silver-Ravenwolf-like" craft names, or to use photographs under a pillow as a spell, or to replace any missing herbs with rosemary, even if I don't personally do those things, even if I wish I knew where some of those ideas originally came from. We learn them because people do them. Even the fake Kabbalah symbols were not just random aimless dreaming by an AI... they are a social experiment that proves an important point about appropriation, that was developed and done by humans.
AI content does all of the shitty parts of this process with none of the redeeming qualities or even childish innocence. And we're being piled with it right now. People are using AI to generate "artwork" that people are buying and using to decorate their Books of Shadows, thinking the symbols are some esoteric magical language when they're just gibberish. People are using AI to generate books specifically trained on the work of actual Pagan and Witch authors without their knowledge or consent, and then publishing them to Amazon to make quick money without doing any real work. Not only are these books plagiarizing actual Pagan and Witch authors, the AI models do not understand these subjects... generative-AI works similarly to autocomplete, stringing together a bunch of words that seem like they'd go together based on the works it was trained on. Guardian article "The stupidity of AI," by James Bridle, says the following:
[AI is] inherently stupid. It has read most of the internet, and it knows what human language is supposed to sound like, but it has no relation to reality whatsoever. It is dreaming sentences that sound about right, and listening to it talk is frankly about as interesting as listening to someone’s dreams. It is very good at producing what sounds like sense, and best of all at producing cliche and banality, which has composed the majority of its diet, but it remains incapable of relating meaningfully to the world as it actually is.
He goes on to say:
The belief in this kind of AI as actually knowledgeable or meaningful is actively dangerous. It risks poisoning the well of collective thought, and of our ability to think at all.
Because generative AI is just advanced autocomplete, and has no consciousness to actually understand what it's spitting out, the text we get from it is famously inaccurate. To greatly oversimply it, it's making things up. Often those things are incidentally accurate, but often they're just not. This is a huge problem, because people who don't understand how generative AI works—who expect that there's some internal computer magic making it accurate—just trust that it is.
As Pagans, we are particularly at risk for this sort of thing, because so much Pagan spirituality relies on unverified personal gnosis, each of our personal collections of knowledge that doesn't come from historical, cultural, or religious research but that we just come up with on our own. For instance, I can tell you that I will use tomato leaves, flowers, roots, and fruit for almost any magical purpose, and that is absolutely correct! It's because I have a personal spiritual relationship with that plant. Most Pagans reading this in context will understand that this does not mean that tomatoes are actually a good plant to use for every magical purpose, it just means that if you work a lot with a particular plant you have a strong affinity with, it will open opportunities for you to do more work involving that plant than just what you'd find in a typical correspondences list. An AI model attempting to relay and reword information based on things I've written may very well spit back that, hey, tomatoes are a good plant to use in magic for literally anything, without any of that needed context, because that AI is not experiencing anything and has in fact worked with zero plants at all.
Imagine that, but instead of just one little slip-up, it's what the entire book is built on. That's what these AI generated books are, and if we aren't vigilant, it will completely scramble up Pagan and Witch knowledge and scholarship.
This issue is made even worse by the fact that so many applications utilizing generative AI spit back information in a conversational manner, providing the illusion you are getting this information from a conscious entity. Spiritual people in general are susceptible to applying spiritual consciousness to things like this, especially if we don't understand how they work, and there are already people developing delusions that ChatGPT is a conscious spiritual guide.
The reality is that AI is not good at the things people are using it for. AI has already recommended ant poison sandwiches, bleach cocktails, toothpaste beef pasta, and drinking methanol. It has also suggested foraging for deadly plants and eating them. A 60-year-old man was recently hospitalized because he replaced his table salt with sodium bromide based on what ChatGPT told him. On a less-deadly level, right now Pinterest is full of AI-generated knit and crochet patterns that result in yarn abominations that look nothing like the picture, because AI does not know how to knit or crochet, it just knows how to make something that looks vaguely like a knit or crochet pattern, and there are loads of fake recipes with fake photographs of food that looks interesting and beautiful but does not actually exist. With this kind of complete nonsense already going on, can you trust an AI to make something like an incense blend or herbal remedy for a Pagan book? You cannot, but that's the world we're rapidly moving into right now.
Something that happened while I was writing this: Google decided to completely obliterate the usefullness of its search by adding AI features to it. Immediately it started churning out (sometimes dangerous) bullshit. Google's AI got a lot of training from Reddit, where a lot of responses are sarcastic shitposts, so it was telling people to use white glue to stick cheese to pizza, to stare at the sun for up to 30 minutes, and to smoke 2-3 cigarettes a day while pregnant. If it's this bad for information everyone knows is important, how much quality control do you think it has for magical advice, which by its very nature is not falsifiable? I mean, when I was like 14, I frequented a Pagan forum where I proudly talked about how I made black salt using acrylic paint. People in the forums immediately told me that was a bad idea... in fact, because it was such a bad idea, I got pounced on long-term over it, and by modern standards that post got loads of "engagement." That post was up for many years, and had it been on Reddit, there's a very high chance if anybody asked about black salt for magical purposes their stupid AI answer bot would have been telling people to mix acrylic paint into table salt to make it, all based on one single high-engagement post by a dipshit teenager. Ugh!
So there's this thing I see a lot where people declare that they have never intentionally used generative AI for anything. Whether they're being self-righteous or just confirming that it's possible not to use it, well... I'm not in this category. When this technology was first coming out, before I knew anything about how it worked, before I saw any perspectives about how problematic or inaccurate it is, I was using it for spiritual things, especially to come up with things like sigils, images of my Gods and Goddesses, and divinations. I was treating it as potentially divinely inspired (a problem in its own right). I was doing this because, while I am creatively inclined in a lot of ways, there are also some creative pursuits I struggle with, and I was playing with the idea of using AI to close those gaps. I abandoned this long ago in favor of just, you know... practicing more. But I wanted to mention it, both because I think it's important to remember that it is OK to change your mind about it if you do use generative AI, and because I have some experience in it.
So let's talk about those symbols and images I was generating.
When you generate a "novel" symbol, chant, or image, it's not actually novel. It's set up to provide the illusion that something is being just magically generated from nothing, but it got that information from somewhere, and it just launders out the original sources, so you have no idea what symbols, images, et. al. went into making that sigil, deity image, chant, or whatever else.
I already mentioned the threat of making something like a toxic herbal blend, but what about toxic symbols? Unfortunately, there are a lot of people who develop Pagan and Witch traditions that are white supremacist, transphobic, homophobic, misogynistic, or otherwise bigoted, and white supremacist Pagan traditions in particular have a whole racist iconography that they use. There are a number of Mjolnir pendants out there that have racist symbols that are obvious to other racist Heathens (and those who study them) but subtle to others... what are you going to do if you generate an image of Thor or Mjolnir or Norse iconography that was trained on these symbols? Yeah, incidental mild similarities between symbols certainly happen, but there is a huge difference between using a symbol that you designed based on your own personal gnosis that has an incidental similarity with a racist symbol and using one that was generated explicitly based on one, even if it no longer entirely looks like it anymore.
And if your "magical symbol" isn't something racist, there's a decent chance that it's based on a fictional series. I found that a lot of my "magical symbols" were slight bastardizations of symbols from Dune, and I had no idea because I never got into Dune. Intentionally using inspiration from fictional works because they resonate with you is fine, but do you really want to be using iconography from science fiction franchises in your magic on accident because that's what the shitty AI you used to generate your symbols was trained on? I certainly don't.
Images of the Gods can result in a similar problem. For one, a lot of people are named after Gods and Goddesses, so if you are generating what you think are photorealistic images of a deity, you may very well just be generating the faces of people who have similar names, and you have no idea how those people would feel about having their likenesses used that way. You can be named Sekhmet or Artemis and not be Pagan. My Patron God is named Set or Seth. "Set" is a common word, and "Seth" is straight up just a normal name, it's even found in the Bible... do I really need to be worshiping an image that looks like some unholy mix of celebrities named Seth? Even more obscure variations of his name are now used for other things ("Sutekh" is also the name of a Doctor Who villain). In addition to being a popular Egyptian Goddess, Isis was a relatively common name for a while... but now, the first results you get in a search for Isis are all about the terrorist group. Without knowing the sources, you have no way of knowing if that image of your Patron God involved pictures of a TV villain, space program, or literal terrorists.
I also personally think it's important to keep the element of human inspiration in our depictions of our Gods. When I was naively generating deity images using AI, I had this idea that by removing the human element I was kind of allowing the divine to speak to me through this technology, but it is ultimately spiritually shallow to think of AI this way, because we already know that the results of AI are not divinely inspired, because Christians have already tried this and the result was "Shrimp Jesus" and a bunch of chatbots pretending to literally be him (regular Jesus, not Shrimp Jesus). You are spiritually gambling if you use this stuff to represent your Gods. It would be way better for you to make an image of your God yourself—even if you aren't a particularly good artist—than to try generating one using AI. The ancient world was full of images of the Gods that are not artistically perfect, but they had human and divine inspiration behind them and survived all these years!
To tie all that together, when doing research I found this image (minus my warning of course) on Pinterest looking for a historical statue of Odin. This is not historical—it's a modern piece—but seeing it portrayed as if it were an ancient piece or reproduction of one gave me pause. While it's unclear if the creator of this statue understood this (the link led to an Etsy shop that appears to be gone), the symbol found on it is a Black Sun, which was created in Nazi Germany and is an explicitly Nazi symbol that has no legitimate non-racist history.
You have no way of knowing if the AI was trained on this shit. And while (at least for now) AI is unlikely to be able to really replicate something like a Black Sun, do you want the images of the Gods you use on your shrine to be poisoned with this sort of thing? I know I don't.
A phrase that pops up in my head a lot is this: Generative AI is the aesthetic of modern fascism. This does not mean everybody who uses generative AI is a fascist, but fascists in particular love generative AI. This is because fascists hate most art (i.e. any art that doesn't promote a fascist narrative) and don't respect artists (they don't believe making art is a skill worth making a living income). This is evidenced by the number of AI images conservatives post of a buff- and masculine-looking Donald Trump, perhaps being coaxed on by Jesus himself, or any number of images posted on the official White House Twitter account of things like Studio-Ghibli-style ICE agents arresting crying immigrants. They love this shit because it allows them to create propaganda without the burden of actually paying artists.
While the similarity with fascists is important to note, the disrespect of artists—or at least the disdain for compensating them fairly—is not limited to fascists, and there is a longstanding problem in the Pagan community with art and book theft for this reason. One of the first experiences I had with the Pagan Internet was going to the website of a painter who specialized in images of Goddesses and finding a whole page talking about how Pagans were copying her artwork to their websites without permission because they wanted pretty imagery on their websites but didn't want to either put the effort in to produce it themselves or compensate an artist for it. Keep in mind, we are not talking about art produced by a huge media corporation that has the means to compensate its artists while eating the cost of some plagiarism, this was an independent artist making spiritual artwork that people found very meaningful, but apparently not meaningful enough to compensate her fairly for her labor. Selling artwork and prints of it to individual consumers is how she makes the money she needs to survive, and people were compromising that by making her work available everywhere, frequently without even crediting her or dropping a link.
I know that a lot of people will have excuses for why they actually are entitled to that work ("I can't afford it!" "I don't know how!" "I don't have the aptitude!" up to the extremely arrogant "But this is my Patron God and this art resonates with me therefore I am entitled to it!"), but all of these excuses are only possible because people chronically under-value the work of artists and other creatives, and don't believe their time and labor is worth compensating.
But the thing is, while it's a shitty thing to do, when somebody straight up just copies a piece of art, especially in spaces like Pagan websites, most of the time they are not actually trying to claim credit for the work. As somebody who has done a number of paintings and drawings of my Patron God, I see them pop up on websites sometimes, and while people never seem to remember to credit me (a simple oversight I'm sure), I have yet to see anybody claim that they made the portion of the artwork I made. Even the people who incorporated some of my Set artwork into projects to post as their own are pretty clear that they didn't draw the things they included, they are only claiming the labor they actually did as their own.
AI prompters, by contrast, are claiming the labor that went into making those images as their own labor. While I would never claim all AI prompter types are equal in their cynicism or delusion, many of them give off a vibe that they believe making real art is backwards, that people who don't use AI tools are comparable to early humans painting on cave walls, and that we are obsolete and should be learning to use AI to adapt to modernity. In reality, AI prompting is not artistic labor at all, it is more comparable to commissioning artwork, only instead of compensating somebody for it and saying "I bought this," they are laundering the labor of thousands of non-consenting artists and saying "I made this" instead.
And I know this is a bit of a tangent soapbox, here, but we should not be aspiring to a Jetsons future. The idea that "the future" necessarily means replacing all traditional, handmade labor and skills with robots and push-buttons and diet pills is not only stupid, it would be unsustainable and a shitty future to live in for most people, who want to make creative things and enjoy creative things made by others, not just ask a computer to do it for them.
Because people who use AI to generate "art" don't value the skill, education, and practice it takes to become a competent artist, they also very frequently lack the talent to see major problems in the work they publish. Not all AI images have major tells indicating they are AI, but I really frequently see AI "artists" posting work they're extremely proud of that doesn't make any sense. For example, here is an AI-generated image of a cracked egg. Everything about this image fucking sucks. The shell is not cracked in a natural way. There is a smudge of random yolk on the inside of the shell even though the yolk is not broken. It's an incompetent image that had no business being on DeviantArt's frontpage. It's also telling that this image has a big watermark, is "copyrighted" (currently AI generated images can't be copyrighted), and is for sale.
To be fair, this person is being honest about being an AI prompter (down to the screen name), so while I'd call it incompetent, I wouldn't necessarily say this person is trying to be manipulative. Nonetheless, the reality is that AI's predominant use—both on the Internet and in print—is scamming people. When this was still a pretty new problem, I ordered a set of deer stickers on Amazon. Not only were they clearly AI when I got them, but there were glaring errors, like one of the deer had five legs (in the listing photo, the stickers were arranged in such a way that a less-obvious sticker was obscuring the extra leg, so these people noticed and just didn't bother to do anything about it except hide the evidence in the photo). This is a huge problem with children's content (there's an excellent episode of Behind the Bastards about this centering on dinosaur coloring books, you can also read a version of this on their Substack if you're not a fan of podcasts), and as I alluded to earlier when I mentioned Shrimp Jesus, a lot of content on social media is automatically-generated AI images with bot comments that only exists to game algorithms or generate advertising revenue.
So... what about examples of AI scams that are actually relevant to Pagans?
After I wrote this initial essay, I started seeing ads for so-called "crystal mugs." They were advertised as if they were functional coffee-mugs made with actual carved crystals like amethyst and rose quartz, but were actually made out of resin (plastic) and generally looked like shit. A lot of people who bought these things are New Agers (not necessarily Pagan/Witch but adjacent/overlapping) who likely were expecting something they could do some sort of crystal magic with, only to get a barely-functional, ugly plastic mug with a cheap aluminum insert.
There is also an increasing problem with Pagan and Witch books having AI generated images on their covers. AI-generated images on a cover don't inherently mean the whole book is AI-generated (something we already talked about), but my personal observation has been that AI-generated covers often indicate rushed, low-quality work in a hard-to describe way... the thing is, some of my favorite Pagan books currently available have very simple covers (like the current self-published edition of Nature Spirituality from the Ground Up, which was edited and certainly not rushed). A simple cover with a photograph or stock image certainly comes off as self-published, but in a way that says to me "the cover was not my priority." I acknowledge I'm biased here (as I was already a big fan of this author before the self-published editions of these books came out), but simple covers or even text-only covers are not a dealbreaker for me if the author is doing a good job advertising elsewhere.
AI-generated book covers, by contrast, give me the impression the person who wrote the book (if they even did write it) wanted a cover that looked professional but were too impatient to bother to put in the time, effort, or money to get a good one.
During a discussion about this subject, somebody I won't name who is involved with a very small publisher was defending the use of AI imagery by insisting she helped "over 100 books get published that would have rotted on people's hard drives." When I looked into her and the books this publisher sells, it became evident most of these "100 books" probably should have just stayed on the hard drives. While I didn't buy any of the books, I was able to get access to the first volume of a series using obviously AI-generated covers, and while I have no doubt the text of that particular book was written by a human being, it was frankly a hurredly-written and confusing screed, with no coherent point at all, weirdly fixated on Christians despite ostensibly being about conflicts between Pagans, and absolutely rife with spelling and formatting issues. There's a point he uses the word CAN"T, indicating he clearly held down the shift key to write a sentence in all-caps, then never bothered to get it edited or even look at the sentence himself ever again. Parts of the book drift between Times New Roman and Arial, as if he perhaps copied and pasted between documents, there are sentences split in half with the second paragraph formatted as if it starts a new paragraph. It seems incomplete. It just isn't a good book, I'd even go so far as to call significant parts of it personally offensive.
That doesn't mean all books with AI-generated covers are equally bad, but it's worth unpacking this a bit, because it reflects a big problem with the way hardcore AI defenders like this person think, which is that they place too much value on quantity (through creating more content or pumping it out faster) without considering how that affects quality. OK, so your ugly AI-generated covers helped "over 100 books" get published... were any of them good? Because if they're anything like the example I read, there should have been no rush at all to get covers created, because the authors should have been taking much more time to edit, format, or even just fucking finish their books. There is no net positive to getting a bunch of half-assed garbage books posted as fast as possible.
Then there's Wicca Magazine. I have been getting a lot of advertisements for them, and this one in particular caught my eye:
I'm almost entirely sure this is an AI-generated image. The mismatched earrings, the mismatched eyes the weird background flowers that don't make sense, random stylistic mismatches, the overall "vibe," and of course this fucked up butterfly are all big tells:
I still periodically get ads for this magazine. So far the worst offender is this one, which features a May pole in the middle of a fire pit, random bonfires at the feet of standing stones, and a string of lights going from a tree straight to the fucking moon:
This is why a lot of people have compared AI to a fae trap... it's stuff that looks enticing at first glance, but the longer you look at it, the more suspicious it gets, like it's the Fae trying to lead you into another dimension and kidnap you or something.
Honestly, whoever is responsible for that should be ashamed, because there are a lot of Pagan artists out there who could use work, and yet they've decided to generate these monstrosities out of plagiarized work instead, without even so much as double-checking to make sure it doesn't have catstrophic errors in the imagery. Furthermore, if a magazine is going to use such incompetent AI on their cover, what will they do to fill the pages of that magazine? If I can't trust a publication to have legitimate cover art, why would I trust what's inside? Was this ever a legitimate magazine? Some of my friends subscribed to it, so I'm not saying the magazine itself is not legitimate outside of its use of AI imagery in its advertising and cover, but I'm not even sure, and I simply cannot trust anybody who uses AI for one thing not to use it for the things that matter.
The main conclusion of this section is that, frankly, using generative AI is frequently a sign that you don't care about the subject and don't take pride in your work, or that there are some parts of that work you don't respect enough to do right. One particularly egregious example can be found in my August 2024 update (a full year ago, yikes), in which I talk about how the creators of "The Transperience," after being criticized for lack of racial diversity in their work, opted to sloppily generate a bunch of images of fake diversity to inject into the project rather than actually involve more people of color. The vibe I got—whether it represents their actual views or not—was that the creators don't care about racial diversity, at least not enough to put real effort into it.
This is an updated section for August 2025, and for me it's I guess a hopeful update, because over the past several months I've noticed this subject being pretty decisively rejected in the Pagan spaces I tend to frequent, and while the reasons for this echo what I've already said, I think it's worth mentioning again: Pagans in general have been finding it invasive and rude when people use generative AI on their behalf without their consent.
For one, people are getting increasingly annoyed when people reply to their questions with "I asked ChatGPT, and..." This is true everywhere—it's like the modern day equivalent of passive-aggressively directing somebody to Let Me Google That For You, it's like saying "you're an idiot for asking us when you could have just asked your phone"—but it's doubly true in Pagan communities because so many of the questions are implicitly looking for personal experiences. Somebody asking whether or not the bottled water they left outside during a full moon on accident is moon water now isn't asking a question with the expectation of a hard-and-fast, "correct" answer so much as asking "has anybody else done this, and if so, what was your experience with it." Trying to get a hard-and-fast answer with ChatGPT is not only not answering their question, but treating them like they should have known better than to ask in the first place.
But I want to relay something that happened over the summer that also really blew up. At the Summer Solstice ritual I attended, we each took a slip of paper with the name of a tree. In the ritual, a group of people representing those trees verbally gave us "prophecies" to mull over. A lot of us didn't remember the prophecies—or we couldn't hear them to begin with—so the organizers of the ritual posted all of them on Facebook. Somebody then took those prophecies and ran them through ChatGPT, proudly posting the answers in a separate post. There was an immediate outcry about this, and the post was deleted, with some discussion about how the group would handle AI in the future. It hadn't been officially banned yet outside of a yearly art contest, but it was I guess vaguely frowned upon?
Anyway, why would people so decisively gang up on a person running tree prophecies through ChatGPT? For one, using ChatGPT misses the point... you're supposed to make those connections yourself, having your own epiphanies, because the meaning of the prophecy you picked is not going to be the same for everybody that picked it, it's going to rely on your specific tradition and life experiences. Attempting to put the answers on a silver platter for everyone using this form of technology that has no consciousness is a misuse of them. As I was writing this, I remembered the outcry when somebody used AI to "finish" Keith Haring's 1989 work Unfinished Painting. While people got really pissed off about it, I believe it's an amazing piece of satire—this person absolutely knows that Unfinished Painting was painted that way intentionally and is not actually unfinished—but the reason it's such fantastic satire is because this is exactly how a lot of generative AI users actually think. Oh, somebody wrote a bunch of cryptic and confusing prophecies that are intentionally vague and therefore open to individual interpretation? Better dumb them down in the dumbing machine so these people understand them better! It was a classic case of missing the point.
But perhaps more importantly, it was wildly disrespectful to the authors of those prophecies, who had written these as sacred words only to have somebody feed them to the AI machine without their knowledge or consent and then had to see the results posted. A lot of people are going through severe AI fatigue, simply because it's so hard to escape this technology... companies have been implementing sloppy AI "features" that can't be opted out of (or are hard to opt out of) whether it makes sense for the products and services they provide or not. Why add to that problem by assuming every question and every Pagan writing should also be fed into that machine?
Obviously, the first thing you should do is, if you currently use AI for any sort of creative thing, be it "art" or chants or sigils or whatever... stop. Again, I can't stress enough, you are not a bad person if you used a widely-available tool without understanding the repercussions (again, I used it for a while, too). But there is simply is no ethical way to use AI for generating artistic works, we have a moral obligation not to use it for those purposes, and it's also vitally important with less creative uses (like asking ChatGPT) that we use our own brains to come up with answers rather than outsourcing our thinking to technology.
But what about other people who either aren't aware of these issues or simply don't care? Let's talk about that a bit.
As a community we are constantly guilt tripped about patronizing Pagan small businesses. I have always had a love-hate relationship with this, because a lot of the things "small Pagan businesses" sell, to be quite blunt, are mass-produced garbage that none of us have any business buying to begin with... these items carry with them the moral miasma of environmental degradation, unfair labor practices, and other questionable ethical issues.
This does not mean that I don't advise supporting Pagan small businesses, I just believe that you should do so in a way that benefits actual, for-real Pagan creators, not just resellers or mass-produced goods. That includes artists and writers. It does not include people who are pretending to be artists by writing a paragraph of descriptions for a plagiarism machine repeatedly until they get a result that almost passes for art. These are more comparable to resellers of mass-produced goods, only they don't just compensate the people who did the real work unfairly, they don't compensate them at all and have the nerve to claim they made it.
So commit to never intentionally supporting AI-generated media in the Pagan community. If a business sells it, or uses it in their advertising, and doesn't stop when they are informed? Don't patronize that business. Don't buy their products. I don't care if they're a small business, if a small business can't operate ethically, it doesn't have to exist, no matter how Witchy the owners are. You absolutely can start patronizing them again if they acknowledge they did something wrong and stop doing it, I'm not in the business of eternal punishment here, but don't just avoid buying AI-generated stuff, avoid buying from people who sell it.
Next, don't share AI-generated works in Pagan forums (or in general, but we're talking about Pagan implications here, after all). If somebody does, point out that it is AI and not ethical. People who knowingly post AI to social media are doing so because it gets them a lot of engagement (which gives them dopamine or money depending on who they are and what social media network we're talking about), so one thing I do on all my social media, Pagan or otherwise, is I hide all posts by people who post AI images, unless I have reason to believe it was an accident. If a forum or group creator is pro-AI? I leave that group. If I'm in a group that isn't run by pro-AI people but regularly gets AI submissions? I hide everybody who posts them (again, unless it's clear they didn't know it was AI). I do not argue with people who are bad-faith AI users. AI-users are largely dealing in a currency of engagement, and angry comments help their ratings just as much as supportive ones. So my personal tactic is to take that away, by either hiding their posts or blocking them, depending on the platform. If you use DeviantArt, there are some guides to blocking and suppressing AI content, such as this one by Stop-Ai-Replace.
If you are an artist, consider using Nightshade and/or Glaze if you post your work online. These are not permanent solutions, as generative AI developers are constantly working to break tools like this, but it will give you some protection for a while, and just for funsies, pro-AI image folks get really mad when people use it. Seriously, some of these people will go on and on about how Nightshade "doesn't work" while also getting really crabby that people use it, it's really funny. You can also use Have I Been Trained? to see if you show up in any AI training datasets and opt out of the ones that honor that. I personally have not stopped posting art online, and don't advocate for doing that (although I am not an artist by trade, I am a hobbyist, and I understand if your opinion on that is different).
Anyway, that's all for now.
Happy Trails,
I wanted to update this to add a couple other examples. One of them is specifically Pagan, the other isn't but it's just such a great example of how sinister AI can be.
The Pagan post had me holding my head earlier today, I found it in a Facebook group for local-ish Pagans. This is a baffling example for me, because it is so clearly AI-generated (specifically the two people on the far right of the image do not make sense... the older man has a shaved head and long hair at the same time in an unnatural way, his hands are too large and extremely unnatural, the person on the far right seems to be melting into the praying man and her lower half is obscured in a way it wouldn't naturally be. While I typically give people the benefit of the doubt (as some people's art styles are easily confused for AI), I simply do not believe that this was not made using AI:
Now, I see AI bullshit in Pagan spaces (and everywhere else) a lot, but haven't felt the need to update this essay. What's different about this one? The combination of the imagery, message, and insistance that it isn't AI is a great example of a particular Pagan irony... the poster of this image wants to "heal the world," but has chosen to express this using an image of imaginary people, generated using a technology that fundamentally makes the world a worse place (including environmentally). They are also pictured in ritual, not actually doing anything to heal the world. Listen: I love ritual! Ritual is great for personal healing and it can be a great accompaniment to other actions when you want to make larger changes. So I'm not going to tell you not to do that, and if this were a picture of, say, a bunch of folks doing an Earth Day ritual or something like that, I'd have no negative things to say about it. But as it has been presented here, it just made me groan, because... simply put, I don't fucking get it. Why did somebody make this post? Is this a joke? I don't even know.
Luckily, people did notice the glaring AI and it looks like people are as annoyed as I am.
But here's a story that is more infuriating than annoying. Last month I went to an art exhibit called "The Transperience," and the centerpiece was a video project with a bunch of interviews of trans people. Great! But as this played out, I started to notice the photo slideshow looked... off. People had weird pupils, which I thought was just an artistic choice at first, but after a while it became evident that a significant number of images in this slideshow and on the accompanying website were AI-generated. A lot of them were high-heel-related, like they'd generated lots of photos of people in shoe stores.
There are a lot of these, and I'm not going to post all of them, but one image in particular made me nauseous, which was this one:
I wound up leaving the exhibit as soon as the film was over, because I was honestly really angry. This exhibit was created by, as far as I can tell, an entirely white group of trans feminine folks, with a more multi-gender group of people giving interviews. As far as I can tell there was only one person of color, a Two Spirit person who talked very openly about that experience. While not as diverse as it could have been, it was not a bad project... except for the blatant, lazy use of generative AI. I very often feel like a crazy person when I'm looking at the AI stuff people post, because it's not just that the use of AI is really obvious, but that it's like nobody bothered to even look at the images before putting them up. I can't imagine seeing that image of the woman with the bright orange high heel and thinking it looked even close to polished enough to go in a video that will be playing at a literal museum, and yet there it is.
And I really tried to be charitable here, in my mind, like maybe they chose these images as a representation of gender dysphoria, but even if that were the case, it's not an excuse for using AI, especially without disclosing that it's AI.
Ultimately, I truly believe that what happened here was that the creators of this exhibit noticed there wasn't a lot of diversity, and rather than attempt to find more people of color to involve in the project, they just generated a bunch of fake people of color as filler instead. And, well... that's simply not how it fucking works. There were a million other ways they could have gotten the filler they needed for this without generative AI.
I don't think I have any extra points here... I guess I just wanted to give some more examples of why AI is deeply problematic and why you should avoid it.
Again, Happy Trails,