I played with A.I. a lot in the process of creating Truth or Dare. My first time doing this with a course.
I think it is enormously importantfor creators and teachers to practice radical transparency in the role of A.I. in their work. So here goes.
Because I am only beginning to discover A.I., a lot of this process is fresh ongoing learning for me. And I want to tell you exactly how I used A.I. to help me create Truth or Dare, and what I am learning in the process.
The first reason is that radical transparency around topics like this is everything to me. If one of my own teachers used A.I. to help create a course, and they hid this fact from me — intentionally or not — that would feel icky to me. My cardinal rule of marketing is, “do unto others as you would have them do unto you,” so I am giving you the honesty that I would want from my teachers.
The second reason is that I think having honest and nuanced discussions about the benefits and dangers of A.I. is the only way we can collectively find our way into responsible use.
A lot of this is a new exploration and ongoing learning for me, so there are no definitive expert answers here. Only a transparent sharing my process and my learning.
That is all we should ever be doing anyways.
So, here’s what I did, and am doing. I used ChatGPT and Claude (paid version), and treated them like an assistant who would help me organize my thoughts, generate ideas, and spit out drafts of what I wanted to create.
I’m pretty cocky about my own intellectual capacity and output. I think I have the best fucking ideas in the world.
So there was no way I was ever going to turn to A.I. for help generating ideas. And whatever ideas it did try to give me sucked. It was nice to confirm that even the best algorithm can’t think anything like I do.
However, where I did find A.I. very useful: helping my ADHD brain get started. For creative ADHD folks, I think sometimes the biggest barrier is that we have these incredible ideas in our heads… and the idea feels so big, the possibilities so infinite, that we kinda get paralyzed. (Raise your hand, ADHD posse….)
When I fed it my big, hairy, ambitious ideas, it spit back out to me simplified breakdowns, summaries, and drafts. This was truly amazing for no other reason than that it reminded my brain that the next step isn’t so complicated, and that I could do it.
Those breakdowns and summaries were rarely usable, but seeing them in front of my eyes closed the gap between the brilliant chaos of my mind and the next functional step. And ADHD folks know: the ability to take the next functional step is EVERYTHING. It is the difference between a finished project and procrastinating for 5 years and being depressed and ashamed.
This did not change my usual creative workflow, but it SHORTENED the time it takes for me get through it. This was, honestly, super fucking cool and something I will continue to experiment with.
Here’s another thing I found.
A.I. wants you to be more concise, have better “flow”, use what I think of as more “standard American” language, remove obscure references, substitute big words with simpler words, and edit your sense of humor to a more mainstream and less edgy one.
But that ain’t me. My natural style is that my thoughts jump from place to place, not always neat linear order. I write run-on sentences, use uncommon expressions and big words, curse a lot, and have a specific sense of humor that isn’t for everyone. (I love poop jokes and sex jokes, what can I say.)
If I allow AI to edit these “different” aspects out of my writing, it no longer sounds or feels like me. At all. And the transmission of the feeling of Simone… is the most important thing to me. Because without that feeling, there is no human-to-human connection. And without that human-to-human connection, there is…. nothing.
So, whatever A.I. wrote, I ended up having to do multiple extensive rounds of edits, adding back the “Simone-ness”, and rewriting sentences the way *I* would say it, and in the sequence and the style that I would say it… to the extent that, by the end, it looked nothing like the original A.I. output at all.
But still, in the end, even after hours and hours of editing and re-writing…. it STILL took me way less time than it normally does. Why?
Because… the “shitty first draft” helped to kick my ass into gear, and into writing mode. And normally I just spend a lot of time in frustrated paralysis.
Once again, I confirm: A.I. is most useful for helping my ADHD brain to get started.
Having a shitty draft in front of me makes me go: “This is wrong, that sounds stupid, ewww i would never do that… let’s get to work fixing this!”
Another pitfall of A.I. is that, since it is so good at spitting out templates and frameworks, it makes it tempting for creators and teachers to instantly produce templates and frameworks for their people.
I could see how easy it would be for people to pump these out ad infinitum, and to dangle these as “more value” for people… especially if they lack creative confidence.
And sometimes, it really can be. I love a good framework… in the right time and place.
But for my course to be an honest reflection of my intellectual ability, I had to ask myself: what is genuinely essential for my people’s transformation? What genuinely comes from my brain, my ethos, the Culture of Me?
What would I produce entirely on my own terms, without A.I.’s. help? And how would I do it?
Anything outside of that, it was clear to me, did not belong in my business. And it was good to come to clarity about this.
I will never, ever have an A.I.-generated template that I pass off as my own. If I point you to a template that did not directly come from my brain, I will tell you so explicitly.
And anyway, if A.I. is so good at creating templates and frameworks that genuinely help people…
… instead of creating them that way and passing them onto your people as if they are your own, you could just tell them which prompts to use to generate them themselves.
In summary:
(1) I’m enjoying using A.I. to help reduce my ADHD overwhelm, sort through the brilliant clutter of my mind, and make it easier for me to see and take the Next Functional Steps,
(2) This is verrrrry different from asking A.I. to do the thinking or the creating for you, and I think we should all be rigorous about not handing over our creative sovereignty to A.I. while still allowing it to help us where it’s effective, and
(3) You will always know exactly what part of my courses are ME, and what has been aided by A.I., and to exactly what extent. I think it is critical, and I think we should all our teachers and creators to the same standard.