Non-Fiction

Encounters with AI

Photo by Igor Omilaev on Unsplash

The images I used in the video accompanying my song wake up were generated by AI. After wasting a considerable amount of time searching for appropriate royalty-free images that fit the lyrics, I resorted to asking Google’s Gemini if it could produce images for each verse of the song. It assured me that it could, so I entered the lyrics. Within seconds nine images appeared on my screen and, with a few modifications, they are what I ended up using. Like most AI-produced images, they were a little weird, and not exactly what I had in mind, but they accomplished what I wanted and their weirdness seemed weirdly appropriate. A bit whimsically, and perhaps with a hint of sincere appreciation, I wrote, Good job, thanks before exiting. Before I could close the tab, my inanimate collaborator spat out the following:

You’re very welcome! Here is a breakdown of how each verse is captured in the individual landscape panels:

Verse 1 (The Sleepwalking Nation): A shadow looms over a crowd fixated on a pyramid of screens, illustrating the “noise and hate” used for control.

Verse 2 (The Fathers’ Labor): An older man looks down at a grave while soldiers carry away symbols of his hard-earned property and home.

Verse 3 (The Mothers’ Grief): A mother clutches her child against a backdrop of industrial fires and marching soldiers, representing the “wars fed by greed”.

Verse 4 (The Patriots): A figure holds a tattered flag of freedom while a giant hand dangles shackles over a divided, shouting crowd.

Verse 5 (The Target Groups): Four distinct panels show diverse groups of people surrounded by torches, highlighting the systematic targeting of the vulnerable.

Verse 6 (The Schemers’ Blight): Tall smokestacks pour pollution into a darkened sky above a poisoned, black river and a barren landscape.

Verse 7 (Corrupted Institutions): A derelict, crumbling church sits across from a cold, modern courthouse, symbolizing the betrayal of the sacred and the law.

Verse 8 (The Zombie World): Decent people walk like hollowed-out “zombies” across a cracked, lifeless earth where justice has died.

Verse 9 (The Children’s Birthright): A lonely child sits in the ruins of history, sifting through the dust of what should have been their inheritance.

I am glad to hear that you find the visual interpretations in the earlier images resonate with the lyrics.

If you are interested in exploring other aspects of these visual narratives—perhaps to adjust the composition, focus on a specific panel, or adjust the mood and color palette while adhering to the same safety standards—please let me know. We can work together to refine these concepts or explore new ways to visually interpret the text while ensuring all content remains safe and appropriate.

I look forward to hearing what we can create next.

I was more impressed with the descriptions of the pictures and the titles it assigned to them than I was with the pictures themselves. Gemini was demonstrating that it (she, he, they?) didn’t just create images that were suggested by the words of the song, but understood the lyrics and interpreted them graphically. It presented that unsolicited information like a human proud of what it accomplished or eager to defend and expound on the work it submitted. I was a little creeped out by its mention of safety standards and ensuring all content remains safe and appropriate, and also by its eagerness to engage with me and collaborate further. I’d heard many stories of people becoming attached to their chatboxes and I had a sudden comprehension of how that happens. As we become more insulated from other people in our dehumanized society, AI agents are poised to step in and fill the void. The dynamic is well-documented in this fascinating piece produced by the NY Times and aired by NPR.

I’ve used AI several times to generate images to accompany Substack posts—the rabbi’s head popping out of a rabbit hole being one obvious instance. I’m reluctant to use AI because I have a lot of reservations about the technology, and feel a little guilty using it. It’s akin to how I feel when I order something from Amazon. I despise Amazon’s business practices, but getting the item elsewhere would be more costly and time consuming—and barely make a difference to Amazon, I rationalize. If I boycott Amazon and all the other companies I dislike, I’d just be making my life more difficult without adversely affecting them, rather like the characters in the old children’s song:

Si and I we went to the circus
Si got hit with a rolling pin
We’ll get even with the goshdarn circus
We’ll buy tickets but we won’t go in.

Si and I we went to the doctor
The doctor gave me an awful pill
We’ll get even with the goshdarn doctor
I’ll take poison and I won’t pay the bill

Whether we want it or not, AI is already a part of our daily lives, lurking in our computers and smart phones, with us when we work, travel, and shop. It’s with us even when we’re not aware of it, enabling the surveillance of our every move and powering the deep fakes which erode our trust in what we see and hear with our own eyes and ears. It may soon eliminate millions of jobs and that prospect is already causing severe anxiety for people fearing the loss of income and purpose in their lives. With such negative impacts imminent, it’s no wonder I feel guilty about any conscious decision to use this technology. I wouldn’t mind the loss of the ability to create AI images if the AI genie could be stuffed back into the bottle.

There is, however, another use of AI that I would miss if it disappeared. It’s something I was hardly aware of when I started using it and now resort to frequently. I’m referring to enhanced Google searches. Yes, I know Google is no less evil than Amazon, but even harder to give up. When using software applications for editing images, recording music, creating videos, and revising my website, there are always things I want to do but don’t know how. Back in the old days (two years ago), I’d do Google searches which would direct me to Youtube tutorials, forums, and various websites related to my request. That was helpful, but I’d have to spend hours watching videos, scrolling through forums full of language I hardly understood, and if I was lucky, I’d finally figure out how to do what I was trying to. These days, through AI, clear, concise, step-by-step directions pop up immediately along with alternative methods. It eliminates many hours of work and lots of frustration, and even changes my attitude about starting on a project that I know will entail difficulties.

It’s not only technical solutions that I find AI useful for but all kinds of inquiries (I assume that before long I’ll be asking AI where I left my keys and save myself some frantic searching). When I came across a book that seemed worth reading, Ben Lerner’s 10:04 (BTW, a brilliant book I highly recommend), I asked Google what the title referred to and received the following response:

The title 10:04 primarily refers to the exact minute in the 1985 film Back to the Future when lightning strikes the Hill Valley courthouse clock tower, generating the gigawatts of electricity Marty McFly needs to travel through time. [1, 2, 3]

The title highlights Ben Lerner’s exploration of time and how past, present, and future fold into one another: [1, 2]

  • Time Travel and Potentiality: The title alludes to the narrator’s desire to exist in multiple temporalities—expanding his own time to deal with a fatal heart diagnosis, a prospective child, and the looming threat of climate change. [1, 2, 3, 4]

  • The “World to Come”: It represents a utopian desire for the future to be “ineffably different” while still recognizable, an idea reinforced by the book’s epigraph referencing the Hasidic belief about the world to come. [1, 2]

  • Cinematic Time: The number also obliquely nods to Christian Marclay’s 24-hour video art installation The Clock, which the narrator watches in the novel. [1, 2]

To explore the critical reception of these themes, check out the in-depth Harvard Review Book Review or the literary analysis from the Sydney Review of Books. [1]

Had it simply answered my query by saying it referred to a critical time in a movie, it would have been sufficient, but as in the case of the pictures it produced for me, AI provided more than I expected, a wealth of useful information along with citations and links to sources. It’s true that it was only able to provide this information as a result of scraping the internet, a practice I deplore, and such easy access to answers poses problems for educational institutions and maybe even for our abilities to research and think on our own, but what a powerful tool it is. The problem isn’t AI, but how we use it, as is the case for gunpowder, atomic energy and, I guess, pretty much all of mankind’s inventions.

Another concern about AI relates not to how it’s (mis)used but the way it’s produced. Here again, the fault is not with the technology itself but with the greed of human beings vying with each other to see who will dominate it (and through it dominate the world). The PBS video below documents the harm being done right now to people, the ecosystem and the environment by the complexes powering AI and the lies and schemes of the powerful entities behind them. It’s likely that inanimate AI has more of a conscience than the unscrupulous maniacs racing to enrich and empower themselves further by controlling it.

To end on a slightly more positive note, I read today that almost half the data centers slated to open this year have been delayed or cancelled. Also, in an amusing twist, AI information is, apparently, becoming less reliable because it is consuming its own AI slop. As an amateur winemaker, I know that fermentation can’t produce a beverage of high alcoholic content because yeast dies off when the concentration of the alcohol it produces gets too high. No organism, they say, can exist in its own waste, and AI growth may, happily, be  self-limiting, a victim of its own success.

June, 2026