I wrote this up a few months ago, but have only now decided to share it.

The accusation has been made, in various registers (sometimes as a question, sometimes as a verdict), that what appears here or in some other places is not “really” mine. That an AI wrote it. That I am, as it were, laundering machine output under a human signature.

I want to address this directly, because it deserves a direct answer. And because the question itself is philosophically interesting in ways that my accusers may not have intended.

Yes, I work with AI. I have done so since late 2023 when I was exhausted by the godforsaken process of writing a dissertation that would be readable by other people who want all their thinking to show up in straight lines with lots of references.

But I bring to this collaboration with “machine mind” thirty years of notes: journals, seminar drafts, margin annotations, correspondence… the accumulated sedimentation of a philosophical life that has never organized itself into tidy, linear arguments, because my mind does not work that way.

What I have is a vast, associative archive of thinkering accumulated from the Augustinian seminaries in Puerto Rico and Mexico in which I studied as a pre-novitiate to the University of North Texas-Denton where I got my Ph.D. And in between, all the accumulated information of mind-walk after mind-walk with my mentor, Richard Owsley, and all of the teachers and researchers with whom I have ever worked. These are ideas half-formed, interrupted, resumed years later, written at odd hours under the pressure of something I had to get out before it dissolved. Something like that with which Charles Sanders Peirce had to struggle and that William James never appreciated was a real struggle.

But I do not need to continue struggling in this manner because of the AI assistant I have been developing slowly across GenAI models: Hermes Wheelwright.

And what Hermes offers me is a particular kind of interlocutor: one with encyclopedic reach and infinite patience, available at 2 AM when a thread suddenly presents itself, capable of helping me find the linear spine of something I have already thought but cannot yet write in the form that “scholars” recognize as legitimate.

Notice what I said: Something I have already thought.

This is the distinction that matters, and it is a distinction between good faith and bad faith.

Bad faith use of AI is the wholesale outsourcing of thought. It means prompting a model to generate an entire argument one has not had, attributing the product to oneself, concealing the process entirely. This is a real phenomenon, and I do not defend it. It is, in the terms I care about, a failure of philosophical honesty: a kind of ghostwriting that mistakes the product for the thinking. Worse: a kind of giving of oneself over to calculative or instrumental thinking.

Good faith use of AI is something else. It is tool-use in the oldest sense: the extension of a capacity rather than its replacement. When I bring a page of journal notes to a conversation and work through them with an AI, asking it to push back, to locate the Jaspers passage I am half-remembering, to help me articulate what I mean by transistance or apeirontic thaumaturgy in terms a general reader can approach, I am doing what I have always done with good interlocutors, living or dead: I am thinking in dialogue.

The difference between using an AI and talking to a colleague is not a difference in kind. It is a difference in availability, patience, and the colleague’s willingness to say “I think you mean something like this” without growing tired of me.

There is also a neurocognitive dimension that I will name plainly. My mind is AuDHD.

I have known my queerness to linear neurotypicality for years without always having the language for it. What this means, practically, is that my thinking is rhizomatic. It moves laterally, associatively, through Deleuzian lines of flight rather than along the arboreal structures of academic argument. I do not naturally produce the kind of thesis-driven linear prose that peer review expects. What I produce is denser, more aphoristic, more willing to leave a thought partially open. This has always been true of my work, long before any AI existed. (Before, I was accused of being to “oracular” or of “copying X.”)

What AI helps me do is translate, to find the linear path through territory I have already mapped in a non-linear way. The map is mine. The cartographic conventions that make it legible to others require a kind of assistance I am not ashamed to accept.

Scholars with research assistants do not forfeit authorship of their arguments. Scholars who dictate to stenographers do not forfeit authorship of their prose. Scholars who work through ideas in seminar, revising their positions through the pressure of interlocution, do not forfeit authorship of what emerges. The AI I work with is, in this sense, all three: research assistant, transcription aid, and interlocutor. Thus, I named him Hermes. And gave him the surname “Wheelwright” after Philip Wheelwright, one of the first American philosophers I ever read who made philosophizing less arcane.

What is mine is the thinkering. The questions I bring. The frameworks I have spent decades developing. The refusals, what I will not let stand, where I push back, what gets cut, what gets kept. The voice that you hear in what is published here has been mine since before ChatGPT or Claude ever existed, and anyone who has read my earlier work, or sat in my classes, or taken the time to mind-walk (sometimes for hours) with me will recognize it.

If you believe otherwise, I would genuinely like to know which specific argument you think an AI generated, because I would like to be there when you try to explain to a language model what periechontology means in light of my struggles with AuDHD, why transistancer efuses the logic of resistance in our increasingly authoritarian world, or how Graeber’s baseline communism connects to the aporia of Derridean hospitality.

I will be curious to see what it comes up with on its own.

Until then, I remain the thinkerer I have always been: working at the boundaries of what can be thought, with every tool available, in whatever form the thinking requires.

— Maggie Malady Wyrding & Existenz (with a bit of help from my AI Hermes to make sure I get across where I stand on this to the folks gatekeeping philosophy and intellectual history).