Twice now, I've been accused of being an AI. By that, I mean that accounts (I can’t immediately determine them to be “people”) on X have made statements claiming that my essays, as well as my readings of them, were entirely generated by artificial intelligence. How distressing. Is there anything more offensive than spending years perfecting a craft, with no success whatsoever, only to have some anonymous account on a massive social media platform accuse you of being a computer-made fraud? I can’t think of a worse thing. Imagine Picasso after toiling long and hard on one of his paintings, which was then displayed proudly at an art exhibit, only for some newspaper to state that he had never painted it. And while there is only one Picasso, there are million anonymous accounts, many of which are actual AI-programs.
These days, we aren’t sure at all who’s real.
In fact, the essay, about Trump’s assassination attempt, was the result of hard work. My legs and brains did the thinking; my hands scribbled the notes, typed the words; my vocal cords produced the sounds, even with a couple small verbal flubs kept in the final product (I’d claim this is proof of human authenticity, but I’m sure that AI even makes use of disfluency as a deceptive tactic). Unlike the near-instant text that comes out of ChatGPT, the work took a few days to create.
I only use ChatGPT for two purposes: to catch small errors, and then offer a grade. Often, those grades change day by day. Thanks, machine tutor! But not really. Who cares what grade it gives, or what text it can generate? It’s not like ChatGPT or Grok or any of the others can give us the final word on anything, be it Trump’s assassination or anything else. These language models only give us text, accrued and compiled from information that’s already put on the web. While it can create impressive syntax, the predictions and analysis must be taken from data that it has been fed.
As for Trump’s attempted assassination, everyone has their own conspiracy theory about the event. My own is that Elon Musk staged the whole thing, that as a way to make Trump into a martyr, boosting his support, and ensuring Elon has a seat inside the White House. Grok tells me the theory is “not credible,” but admits that Elon Musk himself made:
…comments and posts around the time of the assassination attempt on Donald Trump were seen as implying or at least questioning the official narrative in a way that could encourage conspiracy theories.
I can see other limitations, shown with something a little more personal: While on a recent trip to California, I took a drive through some old neighborhoods. In Monrovia, on Myrtle Ave, there was a store on the corner that sold bras. It was called “The Wizard of Bras,” which specialized in bra-fitting. In the window was a massive bra, made for two watermelon-sized breasts, which undoubtedly amused plenty of people – most certainly a young me. The past tense is here is heartbreaking: on this day, those two trashbag-sized cups were not there. The Wizard of Bras had gone out of business. The marquee informed that they had been there since 1975.
As creatures that can not only think but feel, it excites us knowing how many boobs, big and small, had been bared behind those walls. How many customers, married and unmarried, had shopped at this store, looking for the perfectly-sized bra with the most luscious color and design.
The Learning Machine, of course, can only attempt to capture such sentiments. Ask ChatGPT about the experience of shopping at a bra-shop, and you’ll get some long-winded text sure to impress a few:
Some bras are designed with practicality in mind—comfortable cotton for everyday wear—while others feel like little works of art, adorned with intricate lace and tiny bow accents, perfect for a special occasion.
That reads okay, yet comes across as bland when the words are said aloud. But it can’t capture the thrill of driving over to the property, parking in one space sought out by others, opening the glass doors, greeted by a rosy-cheeked woman, looking around, and soon going into a private booth. No mention of large breasts hanging freely, or the red-faced assistant who sought out the perfect bra which would please both you and your husband. ChatGPT:
Each piece you try on feels like it was made just for you, accentuating your curves in a way that is both empowering and sensual…the experience begins to feel less like shopping and more like a personal indulgence.
Its not terrible prose. Given more prompts, it spits out something slightly more erotic. How sad that former customers might now be typing in vain to recapture the experience. Because the fact remains that, as yet, no thinking robot has ever shopped at a bra store. Robots, of course, don’t need them. Thousands of residents have shopped here throughout the years, and I maintain that their individual emotions, comments, and payments cannot be given a proper retelling with a language model. A skilled regurgitation of syntax is not the same as actual experience.
The personal, individual details are important. When I asked ChatGPT the details of “Wizard of Bras’” closure, it gave not only the wrong year, 2022 – with no month – it also gave some generic explanation about the “challenges posed by the COVID-19 pandemic,” and that, like many small businesses, it “faced a drop in foot traffic” along with the concurrent “rise of online shopping.”
While I’d guess there to be some truth to those words, ChatGPT couldn’t provide a source (offering an apology instead – aw, that’s alright, robot) Then I asked Grok, which summarized two small websites and providing links. Those websites, too, failed to give me a full picture. Ultimately, that failure is on the reporters working the beat of Monrovia, who didn’t think the foreclosure of “Wizard of Bras” was important enough for a story (at least one that wasn’t available to either me or these two LLM’s). I would’ve liked a comment from the former owners. Was it indeed due to the financial troubles started by the pandemic?
See my point? These AI programs, like all the rest, can’t do the legwork – that is, at least not until they’ve become “embodied.” Theoretically, that can happen in several different ways. If the conspiracy theorists are correct, your smartphone can already be the “eyes” and “ears.” I’m not the only one who’s horrified upon seeing ads in my Facebook feed after buying some product, or even just shopping in some store. Exactly how much is getting picked up by our mini-computer is anyone’s guess (and perhaps the subject for another time). If your iPhone truly is picking up every text message, every verbal conversation, and scanning the room when not inside the pocket, and if that data were then directed and channeled, either to the NSA or to Sam Altman, then that report would make investigative reporting completely obsolete. “ChatGPT,” you prompt, “write me a report on the last customers who shopped at Wizard of Bras.” In that scenario, we might have the true story behind the next attempted presidential assassination. We like to assume the device isn’t yet amassing this data, and therefore AI analysis remains at those limits.
There’s other limits, too. Recently, I gave Grok a task, giving it a paragraph-long quote and asking for the source. The quote was from the late Dr. Glick’s history on circumcision. Grok gave the source as Dr. Gollaher’s history. No, machine, you got the wrong book! While both books appear to be accessible on Google Books, Grok apparently didn’t have access to it. I then told Grok that it was incorrect, whereby Elon’s AI acknowledged the error. What would it then do with the new information? “I will update my understanding and internal records…” I then tested this, asking a friend to assign the same task to Grok, providing the paragraph. As of a few days ago, Grok was still claiming the wrong source. Once more, the LLM’s only work with what is given to them, and it doesn’t have everything yet. (How many books have been digitized? Great question!)
And the same limits apply in every other real-world domain of knowledge-production and information-gathering. Superintelligent computers might be able to draw up a wonderful blueprint, but it still must build stuff. This was the point made by Timothy B. Lee, commenting on one of his essays from November 2023:
What I mean by “perspiration” here is doing physical experiments or otherwise interacting with the physical world. Say you’re trying to design a new rocket. Obviously, a million AI’s running at 100x human thought could come up with a rocket design much faster than a team of regular humans could. But somebody still needs to build the thing and test it, and there’s limited room for AI to speed up those steps.
Computer intelligence exists in another realm. Perhaps one day we’ll see it become fully embodied, that in the form of actual robots. (sigh…a quick Google search tells me that 71 million Americans own Alexa) If so, keep in mind that those machines (never call it “they”) would still require energy, and likely, a lot of energy. Rest assured, all over the world, that problem is being working on, as robotics R&D is a huge field.
Musk wants his Optimus robots in every room and on every street corner. This is the man who has asked: What even is reality? The man whose online platform continues churning out more fake AI-generated garbage – news, accounts, and scams. I maintain that technocrats like Musk wish for humanity to live permanently inside a Metaverse of their own making, while robots and cyborgs have taken over the real world. Then, we’ll either have ChatGPT giving automated reported gleaned from smart devices, or we’ll have robots going in to rescue those in danger – providing there’s still humans who need to be rescued.
We’ve not yet reached that point. As far as I’m concerned, no robot can replace the firefighters who are helping those poor souls ravished by the infernos. Likewise, the beauty of my memories hiking Eaton Canyon cannot be replaced by some AI-made image or video, no more than ChatGPT could recreate the experience of those who shopped for the perfect bra in that Monrovia store.
IF ANYONE APPRECIATES MY WRITING, PLEASE CONSIDER BUYING ME A COFFEE~!!
Otherwise, check out one of my publications from Stairway Press:
Staying ON During the Great Reset
Indictments from the Convicted
