First, some Generation X perspective
As a solid Gen Xer, I have an advantage as I, like many of my generation, developed a sincere lack of trust in anything authoritative. This distrust of authority started in elementary school where early in my development, I surmised that rules were never imposed for my protection or wellbeing, but only to maintain order and protect the authorities themselves. Growing up in the shadows of the Baby Boomers fed this Draconian interpretation. That ego-driven generation before me pushed every social boundary of acceptable behavior to the breaking point thus evoking an authoritarian push back. Self-centered Baby Boomers so trampled youthful “discretions” that the administrative powers swung the freedom pendulum back like a battle ax against my unsuspecting generation.
It was a very distinct time in history to be a child. All the adults in our lives simultaneously issued constant streams of dictates while at the same time completely leaving us alone to figure it out. Gen X was the least parented generation. Television, save Saturday morning and for a few hours after school, was boring and useless. The internet was still a figment of Al Gore's imagination, and we marauded about town on our banana bikes from morning until the streetlights turned on at night, untethered by cell phones. The only thing we knew was that the rules of our absent adults held very little sway over daily survival in the universe we encountered.
As such, we developed our own code of survival. We learned self-regulation, not out of desire, but as a necessity. In unsupervised baseball games, we called our own balls and strikes. When cut, scaped or bruised, we attended to our own wounds. Most importantly, we avoided interacting with any authorities, whether it be a teacher, principal or town constable, as it always brought a stream of unwanted questions like, "how did you get hurt?" When our circumstances did bring us face to face with an authority figure, we learn to shut up, fake compliance and then move on about our business. Because being children of the '70s meant that whoever the random adult barking orders at us was, they would soon disappear (to where? we never knew nor never cared) and we would be once again free to travel.
This is not just me selectively remembering the days of my youth. Pop culture of the 1970s mirrored this same theme. We grew up watching reruns of Star Trak after school and our hero, James T. Kirk, skirted every conceivable authoritative order from Central Command to save his ship from destruction and his crew from annihilation. Further to that point, no millennial or Gen-Z could ever dream up the Kobeyashi Maru scenario because those generations were never conditioned to understand that when given a direct order, there is always an “option B.” We may have suffered the consequences of our alternative choices, but it never occurred to us that we didn't have the option. Hell, every Clint Eastwood film of that decade, from Kelly’s Heros to Dirty Harry and up to Escape From Alcatraz, portrayed the actor in some form of rebellion against authority. As a free-range Gen-X lad, Eastwood was my North Star.
That same Gen X instinct to question authority is exactly why I don’t take AI at face value
Large Language Models (LLMs) are just another version of an authority structure. It is a machine that, when queried, returns authoritative and seemingly undeniable answers to our idiotic questions just like the Professor on Gilligan’s Island. And when we are fed the results of LLM’s procured answers, we act in same bewildered and unquestioning amazement, just like the rest of the castaways on that fictional island.
I have a ChatGPT account. I have also tested ClaudeAI, Grok as well as other AI models. Here is what I have learned. AI is nice. More than nice, AI is extremely accommodating and complimentary. I ask Chat if the idea I have for a LinkedIn post is good, and its response is "Brilliant idea!" Followed by, "would you like me to draft an engaging post for that topic?" It didn't take me long to learn the pattern that AI is simply programed to blow smoke up my skirt. AI is a model program that speaks as an authority but has no power to perform. And having spent my lifetime learning that anyone, or in this case anything, that holds authority over me never has my best interests as its primary motivator. As such, I have zero problems ignoring the obvious and superficial worldview reflection of ChatGPT.
How do I use ChatGPT? The same way I use a rented car. I drive it as fast and as hard as I can to get from point A to point B. When I am done, I turn it in beaten, broken and in need of a good scrubbing. And, just how I know that the marketing saying that "I am the most important customer" plastered about the walls of the rental kiosk are entirely meaningless words as I wait in line for 45 minutes to pick up my downgraded ride because the last mid-sized vehicle on the lot was just issued to the irritated chap in front of me, I know that compliments from a LLM about my prose, my style and literary ideas are equally meaningless. They are just words some programmer coded into the model to keep me engaged with the robot.
It is curious isn't it, that the only quantity of our entire human existence on this earth that is finite is time. Yet, our time is all technology seems to want to exploit. Social media algorithms are designed to keep us on their platforms, doomscrolling our precious time away. LLMs do the same thing, pushing us ever deeper into their morass as a feature, not a bug, of their system development. Since time being the only truly earthly commodity that we fully control, the tech gadgets that were supposed to free up our time by performing the mundane are now specifically programed to capture as much of it as we unwittingly permit.
How do I use my "rental car?" I use ChatGPT as a research engine. It is fast but its results must be verified. If I do not have personal knowledge of the veracity of Chat's claims or its sources, I ignore the results. I use ChatGPT to draft correspondences. The model is fast, efficient, spells and punctuates everything correctly. I use ChatGPT to help expand writing themes and to draft catchy introductions to my posts. And I use ChatGPT to suggest creative images, charts, and graphs to accompany my work. In short, I use ChatGPT to get as quickly to the end point as I can and then move on with my day.
Finally, I am told I could use ChatGPT by uploading an image of a stack of watermelons in my local grocery store, and it will in return tell me which one is the best to buy. I can't go there. Just like my youth, where I was pretty sure my survival was based on my ability to safely navigate the world around me with no guardrails or adult supervision, I cannot cede to Chat my produce purchase selection. Because, when the chips are truly down, no authority figure ever comes to rescue me from a bad decision. I think my future survival will rest on me personally being able to spot the best melons in the bunch.
Most things in life are more enjoyable when you can place your hands on them and feel their significance. LLMs like ChatGPT will never provide that authentic and life affirming sensation. I have never been in love with a rental car, nor do I ever recall anything significant about one. My use of LLMs simply follows that same pattern. Like every rental car I’ve ever turned in, I’ll use AI when I need it—but I’ll never mistake it for something I own, love, or trust.
Great perspective Bruce.