Hey Clay, What If You Didn't (Use ChatGPT As a Life Coach)?
Prelude
I was at a tech conference from June 2 to June 4. I have been working as a technical writer for eleven some years as of Memorial Day, and this was the first time I ever saw how the users actually used my documentation.
But lemme set the stage the properly.
At the end of 2024, I (much like every other corporate worker) had my end of year review and was asked to think about “my goals” for the next year. And this is something that I take earnestly every single. I have a whole slide deck that I have made for the last three years in the running where I do a thorough analysis of everything that I did, everything I want to do, and perhaps most crucially that year: everything I very much do not want to do. And I did not want to have anything to do with Generative AI.
I occupy a very specific niche as a technical writer. I am about four parts writer, three parts educator, two parts IT worker, and one part beleaguered copy editor or QA tester depending on the exact part of the development cycle. But I am first and foremost a writer, and the the fact that I work for a software development branch of a large technological conglomerate is secondary if not tertiary to my list of sensibilities.
So while I was making my slide deck, I dedicated a nontrivial portion of that presentation was “I have no interest in participating in any of the GenAI initiatives.” If you’ve been reading this newsletter or at least have a vague knowledge about me, I won’t get into the weeds about why, but the point is I was encouraged to pitch a research study at this tech conference instead. Now because tech companies are entirely predictably, GenAI became a part of my research, but I accepted this reality largely because this was my singular opportunity to at least get actual empirical data that generated doc are… bad.
Now, the conference proved that the docs are okay at best and utter nonsense at worst and that also a lot of people are willing to put a lot of confidence into thing that they don’t actually know is right or wrong. Which is terrifying, but also not the point of this particular story.
That was just the prelude.
HeyClay
After pretending to be an extrovert for three days, my energy for dinner pretty was basically enough to go down to my hotel’s restaurant and get something to eat the bar. Americana food gets the job done.
The first night, my bartender is chill, friendly, casual. A far cry from my formal business casual interview I conducted with a coworker for several hours in a row. So, when I go back to the bar for the second night, I sit in his section.
He sees that I’m watching the latest episode of Gundam GQuuuuuux, and excitedly asks me if I mind looking at something he’s working on. A drawing.
First fear: Oh god, I hope he’s not using GenAI to make drawing.
Thankfully, that first fear was immediately assuaged. He shows me his first landscape he drew from watching a tutorial on YouTube. He did on an iPad he bought with Procreate. He shows a time lapse. I ask if I can follow him on social media. Love following indie artist. It is here, I learn his name is “Clay” (It’s not actually Clay, but his actual name isn’t that important to the story).
And it was after I follow him on socials that I discover my first fear is not the worst fear. As it turns out, my worst fear is about to manifest when Clay says
“You know how people say you shouldn’t tell ChatGPT everything?”
I feebly nod fearing the next line of dialog.
“I may have told ChatGPT everything. Trying to get my life goals in order and get better at art. And ChatGPT told me that I should focus on 2-3 things and go from there.”
And that is when my heart fell into my stomach.
I didn’t say anything in that moment. I didn’t have the capacity. But I did hear a voice in my head go…
What If You Didn’t?
I’ve spent the last couple days wondering what I should have said.
The hard line: “no artist of import is going to respect you if you go to ChatGPT for creative advice.”
The mentor: “Hey instead of ChatGPT, what if I put you in contact with some artist friends.”
The community builder (the second cousin to the mentor): “Have you thought about working with other local artists?”
The Socratic method: “why do you feel the need to go to a chatbot for advice?”
I know conceptually why it’s appealing, at least in a vacuum. Therapists are expensive and hard to find. You don’t always feel like you can talk to friends and family about the wilder aspirations. Sometimes, you know what you need to do and just need someone to say it back to you. But I’d much rather consult a mirror than a black mirror.
Maybe because I know how the tech work. The scrapping of untold individuals who put effort into learning something so a computer could get worse at math to do a pale facsimile of something that humans were supposed to do.
Automation is supposed to be for the things that are too hard and too dangerous for humanity. A computer can only ever do exactly what it’s told. I think a lot about the IBM slide that made the rounds.
I think the corollary these times is probably “Generative AI has no way to discern factual statements. Therefore, Generative AI must not be used to make any decisions.”
Post-Script
So Clay, I’m sorry the world has left you with the impression that Generative AI actually knows anything. That the tech was pushed onto the world far too fast and far too recklessly that you think it is a valid place to seek consul. I’m sorry I didn’t have the words then to try and course correct. But maybe, next time, I can shape a response in the moment.