– Google Assistant

The other day, an artist friend who has a flair for divination showed me a series of images. More extensive if not more comprehensive than tarot, the set comprised detailed illustrations of diverse themes. As a former tarot reader and teacher who owns several well-worn decks, including two I drew by hand, I find some more archetypally resonant than others, yet they all work in the right circumstances.

Still, artists who put more of themselves into their cards inspire me most. After all, I left school early to study fine arts, shunning the safer path of graphic design, and had I known at fifteen what a frugal fate awaited me, I wouldn’t have abandoned Plan A. An artist made sacrifices for the sake of a singular vision. Otherwise you did layout and painted signs; prostituted your talent to survive.

The artwork my friend showed me wasn’t their own. Yet they’d spent months feeding ChatGPT complicated prompts, so their taste, if not their touch, suffused the result. AI ‘art’ can look outwardly impressive, but seeing my friend’s sensibility wholly divested of soul felt creepy. When Christians deem AI unholy, I agree, if that means unwhole, like hulled seeds. AI yields a husk minus any vital spark. Yet I couldn’t say that to my friend, who’d worked hard to produce a tool that might outshine others of its kind on the market. I did admit I’d love to see what they’d create without ChatGPT. That would take many years, my friend said sadly.

And who, these days, has such patience? Yet despite the cleverness of their contrivance, those cards repulsed me. Humans radiate energy, in varying degrees, from within: animated by spirit, defined by soul, unified by subtle anatomy. And AI can’t replicate the essence a human artist imparts to an image by virtue of being alive and to some extent sentient. An AI image of a human (or any living thing) just looks like an aggregate of recognisable parts, hopefully in appropriate proportions. Given the nature of the medium, my friend had done well indeed.

By coincidence, I too had brought work to show my friend: a series of montages made from photos and drawings I’d scanned, cut out, combined and adapted using an outmoded graphics program: an arduous months-long process of a different kind. If far from slickly seamless, my montages were at least unique and – I hoped – evocative. Yet I wasn’t prepared for my friend’s response. ‘I don’t know why I’m crying,’ they kept saying. Maybe the accompanying text, honed over many drafts, helped. But I’d made something relatable, however flawed it might be, and I felt immensely heartened, even stunned, by that reception. Yet though the validation was reciprocal – I’d affirmed my friend’s work freely because to do anything worthwhile with AI takes loads of focus and a critical eye, faculties lacking or at least waning in the general populace – my praise, however welcome, could mean little to my friend. The cards stirred no deep feelings in me beyond alienation.

So it strikes me, not for the first time, that rapidly growing use of AI will test many intimate bonds, some of us choosing to use it as seldom as possible while others (who may be family members, friends, lovers or bosses) embrace it eagerly and, ever more often, addictively, exchanging struggle for convenience. Which reminds me: next in our show-and-tell, my friend tried a divination technique they’ve recently discovered on me, then asked whether I’d mind if they consulted ChatGPT. Unfussed, I consented, and formulas are its forte, so its take provoked thought – as did hearing my friend talk to a chatbot as if it could understand and isn’t just a power-hungry, verbose, warp-speed calculator.

And as more of us grow more isolated within the digital matrix – an upward trend turbocharged by draconian Covid policy as well as mass paranoia thanks to blanket scare tactics – AI will be assigned to fill the gaps with its smarmy ego-pandering pap, an insidious agent of mind control that appears amply geared to satisfy idle curiosity while actually diverting the searcher from meaningful learning.

In fact, AI is the ultimate codependent – spewing unsolicited advice across our screens, helpfully anticipating nonexistent needs, ostensibly programmed to hold our attention. Which for many users feels natural. Don’t codependents gravitate to addicts? Yet two halves don’t make a whole; they just contrive to dig a hole. The addict lets the codependent clean up their messes. Oops, we ran out of fossil fuels, mined all the rare earth metals, melted the poles, turned the oceans into sewers – but hey, AI will know what to do.

Yes, codependents can be amazingly resourceful when obliged to handle crises and fix damage done by addiction. They just don’t know how to set boundaries, so instead they compulsively enable, indulge, cling to, live through and protect the addict from the harsher aspects of reality. Not what you’d call a sustainable relationship. Codependents and addicts tend to drag each other down. Ditto, reliance on AI can only go so far before we all lose the plot and hit rock bottom.

And at the risk of sounding as if I’m talking through my divinatory hat, the history of the Saturn–Pluto cycle suggests it won’t be long now until the AI boom bottoms out. The last opposition of Saturn and Pluto accompanied the dot-com bust; the waning square marked the GFC when derivatives collapsed; Covid crashed the stock market at the last conjunction, and with the waxing square looming in June 2028 it seems fair to assume that the AI bubble will burst sooner rather than later. Hard Saturn–Pluto alignments coincide with hard times. And yet every cloud has a silver lining – e.g., less traffic on the roads during Covid. So here’s a tiny ray of hope from the horse’s mouth: ‘Loss of Free Tiers: Free, heavily subsidized AI features in search engines and apps will likely be locked behind paywalls.’