I give up on not writing about AI.

I've talked about my frustration about being flattened into one persona on the internet before, and I've talked more recently about the struggle of privacy when it comes to single, connected profiles. But there's one area in which people right now are particularly fast to eliminate nuance, and that is in regard to AI.

I wear many hats in my life, and in each role, my relationship with artificial intelligence is very different.

In my role as a graduate student studying Philosophy, my interest has always been how people categorize and come to understand information. I want to understand why we think the way we do and what shapes those thoughts. The arrival of LLMs and generative AI felt like divine timing for my grad student journey. My admissions essay compared generative AI to Plato's Allegory of the Cave. Several papers throughout my academic career have touched on different aspects of AI and how its outputs, flaws and all, are mirrors of issues faced elsewhere by humanity. It is a part of a larger topic I find fascinating and I want to talk about it.

Because I've been studying these things, I felt it best to learn to use generative AI, enough to understand it firsthand. And to be clear, I've tried a range of models and harnesses in pursuit of finding the most environmentally friendly and socially conscious ways to engage with AI academically. I use local models where I can. I don't ask for information I can't obtain or research through easier means. I do not use AI to replace my own thinking, and I always read AI generated results as one perspective that might be flawed or biased.

While the hats I wear may be different areas of my life, that doesn't mean they don't influence each other.

A year or two ago, my day job started experimenting with adding AI to our workflows, as organizations right now are prone to doing. We are expected to try AI, to share our successes and failures. As such, my coworkers noticed I had experience with AI beyond just "write prompt, validate output." I can diagnose prompting issues and speak to the technical side of things in plain language. And as such, I have been asked several times to test agents or weigh in on documentation that is being handed to AI.

I try to use what I know to help my colleagues, in the hopes I might help. I share what rules I've made for myself when using AI, to help others set their own boundaries. Most importantly, I point out that using AI shouldn't remove the joy of one's job, just alleviate the things causing burn-out or frustration. We make do in this new "AI-augmentation encouraged" landscape. 1

Inevitably, I have found use cases for AI in my own life.

I was diagnosed with ADHD as an adult, answering years of questions I had about myself and my 'bad habits' with one external answer. It's not me, the mind inside the body; it's the body chemistry and wiring that is influencing the mind.

Part of coming to terms with my diagnosis was finding new methods of handling my executive dysfunction. I don't call it 'battling' anymore; I fought that fight for years and always came out the loser, either from pushing through at the cost of burnout or the shame spiral of failing. I must have tried every task/project/life management app on the planet to try to get myself organized, and every one fell by the wayside because, the instant I set it down, there was too much 'weight' involved in picking it back up. I entered boom and bust cycles, where tasks and habits got lost when I transferred between systems. I found the things I liked and that worked, but they always included upkeep that I felt I couldn't rely on myself for.

But AI could do that. From what I learned from my academic pursuits, I had the boundaries I wanted for what AI could and could not help me with. And with work insisting I "try using AI," showing an elaborate task management system is a fast visual to check that box.

I have to admit, AI has helped me with executive dysfunction immensely. I'll save the 'how's for a later time, but the point is what I've created using Obsidian, a slew of markdown files, and a bunch of scheduled AI processing is an individualized solution that no AI-less app has managed for me before. And it is such a relief to have it as part of my life.

But there are parts of my life I will never allow AI to have part of.

It says it in big bold letters in my vault's instructions to any agent: any sort of generative or agentic work regarding my book series is off-limits. My writing vault isn't even connected to the one I let AI work inside of. At most, the agent will know when a chunk of my time in a week went to writing and might congratulate me on my progress. It cannot see the words.

Writing is a craft. It is more than just what it produces, and it is that process of production that will (hopefully) create an end result I want to share with others. It is the highlight of my day to sit down and write. My co-author and I will spend hours discussing and negotiating scenes. Why would I give away any part of that to an AI to do for me?

No, AI has no place there. My writing is my own creation.

And yet, as a human being, I cannot perfectly segment myself.

The demands of life and switching contexts mean that I am constantly using the same hardware to work on things in multiple contexts. I use a variety of browsers and search engines for my day job, meaning I don't always remember what browsers have AI search results turned off. I'd be lying if I said I'd never looked at them.

And while I do prefer the assistance of an academic librarian when I am struggling to search for something for class, having AI perform that role in a pinch can close a gap in my knowledge, especially when I am combining concepts from multiple disciplines I have no prior experience in to find a plausible answer. I write speculative fiction. I cannot study all of communication theory to know how a government would write a press release after a Godzilla attack. I don't know how to find out if someone resurrected by magic would have bad breath because of the bacteria running wild in their mouth. AI gives me a starting point for researching on my own.

But does that mean it wrote for me? From where I stand, no. The words are still mine, the ideas still mine. Not everyone is going to share the same position.

The irony of this when working on a series about moral grey zones is not lost on me, though I kinda wish it was.

I fear what people will say when they see I am a multi-faceted person.

We live in a time where acknowledgement of using AI in any way is a steep and slippery slope to stand on. I have avoided talking about AI because when my co-author and I go to publish, I don't want my academic interests to drag my creative passions down as a result. I fear that, by virtue of reading so many AI produced texts in my daily work, I will inadvertently mimic it. Is my understanding of an em-dash, courtesy of my graduate studies, going to come off as AI written to a casual reader? What can I do to head this off?

I have seen the rise and fall of many technologies over the years. Over time, conventions are formed and norms are established. But I cannot - will not - let the lack of conventions around AI now hurt the writing I love in the future. But at the same time, I don't want to let it stop me from sharing my thoughts on things now in the realm of AI.

Publishing things, making them public, especially things close to the heart is always a risk. It is always an act of courage to do so. I want to be multi-faceted, to show who I am in all my passions. I am filled with hope for the future at the same time as I feel the overwhelming pressures of the world pressing against those same hopes.