I wrote my last post using AI.

A few months ago, my friend Akemi and I sat down to our bi-weekly writing session. Our usual process is: sit down for one hour, and see what happens. That week was my birthday. For the first time, I was asking my close loved ones to spend time with me, doing things I liked. I was wondering why this felt like such a big ask. I remembered how, as a kid, my parents would beg me to not do anything for my birthday, because it was a lot of work. I started to write about that.

Akemi and I paused at the half-way mark to read what we’d written. When I read mine back, I felt bored and irritated. It didn’t feel profound or self-aware or smart or kind. Or at the very least, useful. Just another observation about how my parents sometimes made me feel like a burden. Hadn’t I already written about that? Hadn’t I already done healing work on this? Read all the books and done all the therapeutic practices recommended by science and the New Age? Was I ever learning anything at all? Did I contain any creative expressions beyond my own pain? Beyond centering myself?

That was when I decided to stop feeling dumb and boring, by outsourcing the task of sense-making to the machine. I switched over to AI, and asked it to write a substack post which was “mythic,” “surreal,” “uncanny,” “creaturely,” “emotionally resonant without self-disclosure,” and which “obscures meaning intentionally,” reveling in “the power of being hard to comprehend.” This set of words had itself been generated by the AI a few months prior, when I asked it to summarize my feedback on a poem.

For the content, I asked the AI to use the exchange we had had a few days prior, during which I shared some relational observations and theories, wondering where I wanted to stand in the spectrum of easeful (“river”) to activating (“forge”), and for which relationships.

The AI spit out a few options. I selected the one I liked best, then spent twenty minutes editing it to my tastes.

I read it to Akemi at the end of our hour, pretending I had written it. Her response was pensive and affirmative—pondering her own perspectives on the content. I felt a twinge of shame for my deception. And futility… What was even the point of writing if I was going to lie? This substack has always been a self-serving exercise in personal courage: allowing work to be seen, even when I do not feel it is good or done; experiencing “completion” as a conceptual construction, the bounds of which I can decide. Coming from academia, where my intellectual projects don’t feel “complete” after years of work, I wanted to practice taking the risk of exposure (clicking “publish”) early and often.

So, I had just wasted my time pretending to create. Dismissed my own capacity to channel my life force energy into manifesting, like it was incompetent and worthless. I might as well have been miming typing on a toy computer.

That night, I shared the story with a loved one. I found myself virtuously adding how excited I was to write all about it in next week’s post; all about the moment at which I decided my own creative energy was not good enough, and I handed the task to the AI. He affirmed my authenticity and realness and wisdom in making the admission and examining it. I knew that as long as I admitted to it and shared a deep and self-aware reflection the next week, then the unconsciousness and cowardice and insecurity and deception would not have been for nothing. My integrity would be restored. I’d named my substack “iamstrangeloops” to invite me to incorporate the infinite loops of reflexivity that usually paralyzed me into the writing. Here was an opportunity with a high authenticity-pay-off.

The next session, I sat at my desk for one hour, and tried to write what happened. What came out was the post about my parents and birthdays which I had meant to write before. I felt all the same dissatisfaction, concluding it was self-centered and repetitive. So I didn’t post it.

Two weeks later, I tried again. What came out was a cleaner version of the birthday story, plus some grandiose reflection on AI.

I skipped the third session.

I felt all the same feelings that the things I make are not worthy of my attention.

So here I am a fourth time, trying to say something true.

I don’t have the answer, for how/if to value the creations which are unconscious, small, trite, incoherent, and human.

I know the next right thing to click “publish.”

My hour is up.