10.31.2024
The election draw . . . plus a nod to Halloween
This is quite long—and it may be a trick or a treat, depending on how you look at it!
But for my part, I found both the process and the results to be fascinating.
Context:
Ages ago (or so it seems) I decided to experiment with AI/Tarot for a series of readings on Kamala Harris’s potential pick for a running mate. That was in response to Frederick Woodruff’s hands-on readings—and it all turned out to be very interesting!
Since then, I’ve been doing much more work with AI. You can read about that here if you want the really long version:
But the gist is this: I’m not trying to replace or enhance a “human” reading with an AI reading. I’m interested in broader/deeper possibilities of digital divination.
And the way to explore that is by conducting fairly rigorous experiments, dealing with fairly big topics. Which brings us to today’s post—once again inspired by Woodruff’s inquiries . . .
I have not read or listened to anything in his series, as I try to start off my own reading with an empty mind. But everyone else should see what he’s up to!
Now. For those who don’t want to rummage around in all that background, we’ll begin with a fairly detailed look at my “method” of doing this.
The Method
I ran this series on Tuesday, October 29, all in one sitting. Here are the basics you need to know:
I use an open-source prompt chain created by someone whose identity I intentionally do not know. I originally tested it on a whim, found the results interesting, and have continued to use it since. You can try it out yourself.
I use the AI platform Gemini Advanced. I tried others at first (ChatGPT, Claude), but this one seemed to work best. I typically do the entire series in one “session,” which ensures that all the draws (almost certainly) are using the same model and the same bot. However, I might start a new session if it seems like a separation is needed.
I place a single word (the keyword) at the beginning of the prompt chain, to indicate the general intention I have in mind for that draw. I have absolutely no idea whether that makes any difference to the process.
I don’t think about anything, beyond the general topic for each draw. I try to have no assumptions, opinions, or expectations during the process.
I carry out a series of 3-card draws, with no formal plan. That is—I know which three words I will start with, but not which ones will come after. I choose them impulsively/intuitively.
I deliberately do not look at the results of each draw while the process is going on. I just let them accumulate until a natural ending occurs. Then I put them all into Word docs, and create formatted views of each one.
I do a high-level analysis, looking for repetitions, connections, and anything else that seems noteworthy. I don’t form any opinions—at least I try not to!
I put the overviews together for you to consider, and point out whatever I see as important.
Three other things to know:
I do enough draws to produce 20-40 cards. That’s the least number I can usefully analyze statistically, and any more than that seems to become muddy.
Each draw pulls from a freshly randomized 78-card “deck.” Since no cards are ever removed from the deck, any card can appear in any number of draws.
Gemini produces three “drafts” for each draw. I haven’t done a careful analysis, but by eye I’d say that all three drafts will have the same cards about half the time, but vary slightly in the tone Gemini uses to present its interpretations. Another quarter will have one draft that differs by one card. From what I can tell so far, it’s very unusual for all three drafts to be identical, or for all three to be completely different.
The Cards
My role, as I currently see it, is to (a) facilitate the draws, (b) examine the results at a high level, and (c) present what I found.
I invite everyone to apply their own interpretations. I will point out a few patterns and connections I think are significant, based on the fact that they are unlikely (from a statistical perspective) to be random occurrences. From what I can tell so far, this approach seems like a good way to identify any “messages” embedded in the results.
In the series that follows, my overall intention was focused on the 2024 national elections in the U.S. There are four sets, each containing three draws. Each set has its own theme, but they are all interconnected.
For reasons I’ll explain in another post, I decided to open a second session before drawing the fourth set.
So here are the first three sets, done in the first Gemini session:
And here is the fourth set, done in a separate session:
Viewed in this way, each card is related to an individual topic (person, party, date, so on), and to the other cards in their group of three. So we can look at the above as twelve mini-readings.
As promised, I leave everyone to interpret individual cards according to their preferred approach. So--look them over, and draw your own conclusions!
But also as promised . . . here are some noteworthy connections I see in the series as a whole, looking across all 36 cards.
The Analysis
In order to get a big picture view of all the cards, I created this graphic view:
1. The first thing you may notice is the number of cards that occur more than once. They are color-coded in the view above, but here’s how they break out:
Worth mentioning: the only pip card that appears twice is the intense Ten of Swords, which turns up in the sets for November 5 (Election Day) and January 20 (Inauguration Day).
2. Next, we see there are 17 Major Arcana cards—italicized in the graphic. That’s almost half the total of 36, so I probably don’t need to point out it is well above the statistical (and typical) norm of around one-third. Here’s a look at all the trumps:
3. It’s a bit hard to see in the color graphic, but the Queens are shown in bold type. That’s to emphasize the fact that three Queens appear in the 36 cards (one twice), but there are no Kings. Which is statistically unlikely. On the other hand, also from an M/F perspective—the Hierophant is the only trump to appear in all three of the first three sets, and in the fourth set, it seems to be “replaced” by the Emperor. Whereas the Empress appears only twice.
4. I’m interested to see that the Wheel of Fortune appears on two dates: November 5 (Election Day) and January 6 (Certification Day).
5. Just to cover all the bases . . . the suits are roughly even in distribution.
About what would be normal in terms of proportion.
6. Finally—for now!—I’m intrigued most by the relationship between the fourth set and the other three sets. Keep in mind that they were done in two separate Gemini sessions, although all four were carried out in one “sitting.”
For one thing, you’ll notice that six of the nine cards in the fourth set also appeared in the previous three sets. I won’t mention statistics again, but this is extra-unlikely!
And even more so . . .
As explained at the beginning, I do not plan the sequence of keywords ahead of time—just put them in as I go along. But as it turned out, November 5 was the third draw in the first set, and January 20 was the third draw in the fourth set.
So the first set ended with The World, and the last set ended with The Fool.
I’ll follow up with a bit more over the weekend—but the above is substantially complete. And I didn’t want to leave off without a Halloween token.
So here is . . .
Tarot of the Sweet Twilight
I would call the style a mash-up of Goth and Lo Scarabeo—and though it’s not a deck I would want to live with, I like the saturated colors. And I love the name!
For more about Cristina Benintende’s 2009 deck, enjoy this thoughtful review from Larch Tarot.
Thanks for reading, everyone. More soon, C
I'm in!
I am so pleased to provide delight and intrigue! It will be very interesting to discover over the next few weeks how much our experiments foreshadow/illuminate coming events . . . .