A Little Note on Using Generative AI

Perhaps, it’s not always pragmatic to be a person

Generated with ChatGPT

“Sometimes [the Bishop] dug the soil in his yard, sometimes he read or wrote. For both these kinds of work he had just one word: he called this ‘gardening.’ ‘The mind is a garden,’ he used to say.”

- Victor Hugo, Les Miserables

When Buzzsprout—my podcast hosting platform—launched their Co-host AI update, I was thrilled. I signed up for the added $10 a month, excited by the prospect of reducing my production time. Each podcast episode takes around three hours for me to produce and post…and that’s not including the scheduling of the guest, the background research and reading, or even recording the interview itself. When everything is factored in, it probably takes a conservative average of fifteen hours per thirty-minute show.

For me, my favorite part of running Aiming for the Moon is the hosting part. I love preparing for the interviews. I love reading and thinking and discussing the ideas of my guests. I’m captivated by our conversations. For a while, I have wondered (and dreamed) of having an editor for the show to streamline this process. Someone I could trust to edit the audio together, type up the descriptions, track down the episode topics, and (my least favorite part of all) upload the episode to social media.

So, when Buzzsprout announced their Co-host AI update, I was excited. Perhaps, now I could off-load some of the administrative tasks to this new program! Its abilities to write descriptions, suggest titles, and mark chapters could significantly reduce the time it takes me to release episodes. And if these nitty-gritty details were handled, I could do more interviews!

After signing up for the upgraded plan, I was amazed by the efficiency of the generation. And yet, something felt off. As I continued experimenting, I realized that I could shorten the AI-created descriptions into intros which would speed up the process even more.

Despite my excitement, there remained this odd, sheepish feeling. A smile with a twinge of guilt. It felt odd to hand my work over to a machine. To give it my job of writing. It felt like I was stealing or cheating the artisans’ system.

Unsettled by the ease, I returned to writing my own intros. I realized that the speed might compromise my practice. The painful, annoying, tedious process of turning my thoughts and impressions into letters, words, and sentences was part of the point.

I realized that I needed to craft my own intros because I need to practice. If I gave my pen to a machine—even temporarily—I might get used to the ease and forget how. Don’t be swayed into neglecting the “soul-leading” that is writing.¹

How was I to become a better writer if I never wrote?

“Writing isn’t hard work, it’s a nightmare” - Phillip Roth.

It’s painful to reach for an idea and try to formulate it into words. To try to cram it into syllables, distill it down into letters. It can be an agony to capture my fleeting thoughts in Times New Roman. Typing ideas cements them, in a way, to the page.

But, what if part of the beauty is the struggle?

We have a tendency to rationalize all life and reduce it down to tasks. The “virtue of speed”—as Prof. Adam Smythe (ep. 126) put it—has commandeered the hearts of humanity. In praise of production quantity, we have shortened the process.

I am not an AI-hater by any means. In fact, I plan to study computer science with an emphasis in AI for the next four years. I am captivated by the concept of math digitalizing patterns. But, to me, uncovering the math beneath poetry doesn’t reveal that prose is the sum of its parts; it deepens the mystery.

What if part of the magic is lost when we cut out the painstaking process? What if part of the glory of the produce was the blood, sweat, and tears that watered the garden? Speed, systems, and symbols are wonderful things.

But, what happened to waiting?

To patience?

To silence?

If the field of computer science is built upon problem-solving, perhaps we just need to think deeply about the “problems” we are trying to solve.

Intelligent use of AI is not relinquishing your mind to a machine or avoiding the algorithm altogether. It’s about being intentional with automation. LLMs (like ChatGPT) are allowing us to personalize automation. I’m not writing to instill fear of AI—to manifest Frankensteins in the costumes of chatbots. But, if AI is to act like a puppet, let’s be clear who’s pulling the strings.

Don’t sacrifice your soul to silicon. Each of us has the ability be creative. We have the ability to think, to wonder, to dream. The way we apply these tools in our own lives will be different. I use generative AI for graphics for my Substack. Why? Because my art is my pen. For me, having a machine generate text is sacrificing a deep part of myself.

Perhaps, it’s not always pragmatic to be a person. Perhaps, our mantra should be not to swear off tech, but to think about it. Use it! But know, how we use technology changes us.

What are you sacrificing when you press “generate?”

“To be lost in thought is not to be idle. There is visible work and invisible work.

To contemplate is to toil, to think is to do. Folded arms are busy, joined hands are active. Gazing heavenward is industrious.”

- Victor Hugo, Les Miserables

If you enjoyed this, you may be interested in my article on how AI’s imitation of art reveals the uniqueness of humanity:

1. Plato’s Phaedrus 261a.