# AI or LLM Usage Policy There are a lot of reasons why people may want to know my particular stance on Large Language Model usage (often called AI). Some people may want to know because they're curious or nosy, an impulse that I understand and share, while others may have a moral, ethical, or professional interest. This page is for all of you. My general stance on LLM technology is one of deliberately limited engagement. My passion and profession, really my vocation, is based on the written word and the creativity of the people who write, particularly on the topic of the things of God. The few, specific instances where I am comfortable using an LLM chatbot or a neural network tool have almost nothing to do with the process of reading, writing, synthesizing, ideating, engaging with such works. Let's get in to specifics. ## Positives and Negatives I want to describe the specific jobs I will use a LLM tool for in my daily life. + I occasionally use a neural network tool called Whisper that openAI developed and released open-source a while back. It accurately transcribes spoken language audio, like podcasts or audiobooks. It runs locally on a PC I also play video games on. I used it quite a bit as a student to transcribe my own first drafts for papers I had a difficult time getting on the page, and these days I keep it in my back pocket as part of my bag of tricks. + I occasionally use an LLM chatbot to search the web for information that is techinical in nature and kind of hidden inside parts of the web that are obtuse or difficult to navigate, like databases of owners manuals for VCRs from the early 2000s. I do this for some hobby projects of mine that are related to retro computing or retro gaming. A real life example, I asked a chatbot to search manuals and forums to see what the maximum size of memory stick a PDA I bought would support. Next, I want to describe the specific jobs I will not use an LLM for. - I don't use an LLM to write or draft at all. This is the muscle that needs to be developed over time in my profession, and I remember what being a novice feels like. I can't outsource my drafting to a bot for the same reason I can't outsource it to a research assistant, I need to keep developing and growing as a writer for as long as I can if my work will ever approach the work that inspired me to take up this vocation. - I don't use an LLM to research or break down material at all. In the past, as a student, I experimented with limited use of an LLM tool to generate an outline of a difficult source that I could use as something of a roadmap while reading. It was useful and about 90% accurate. It wasn't accurate enough for my purposes as a student. My standards are higher now, not lower. See also the above, if I'm going to continue to grow as a writer in my field, I need to continually grow in my ability to digest sources and can't afford to let opportunities to practice go by. - I don't use an LLM to do human work. I am not a visual artist. In the past, I have experimented with using a tool to generate visual art for a profile picture, for an article header, etc. I decided that, beyond not being impressed long term with the results, this represented a pretty serious solidarity issue. If it's work that I think should be done by humans for human reasons, I don't use LLM tools to undermine those humans. Better by far to mine the public domain for something suitable, if you can't afford a professional. ## Reasons Why Some people are very interested in the reasons why, and I approach this section with some trepidation. I run in circles that are no stranger to something I call an activist mindset, the sort of worldview that is constantly engaged in a war of language with an opposing side, generally with existential consequences on the line. Frankly, there are people who will run straight to this section to find the one way we don't agree and try and push my perspective towards theirs. They live to argue. They live to post. I'm not generally amenable to individuals with Poster's Madness. This section is really written for people who are curious as to why I limit my LLM usage in the year of our Lord, 2026, when it seems like there are many pressures towards LLM maximalism. 1. As a writer and a thinker, LLM tools are specifically designed to replace my labor. I think they are a poor replacement, but if I am a laborer, they are a scab. 2. As a human being, I learn by doing and repetition. If I'm going to get better as a writer and thinker, I can't use these tools for any meaningful work. 3. The environmental cost of training these models is disasterous, which is difficult to ignore. We are well past the point of diminishing returns on the power and physical infrastructure spend going on, and in a functional democracy we would have put a stop to it a long time ago. 4. A more subtle point, the biases that are baked in to LLM functioning and output are antithetical to my beliefs about reality. This point may be worth expanding on. ## Reality The work of [[Charles Taylor]] and [[Andrew Root]] has been very formative for my and my conception of the modern values of [[efficiency]] and [[speed]]. Taylor describes these two goals as main features of [[secularism]] and [[modernity]], and that their pursuit is often inhumane and creates twisted versions of humans that seek to center those values in their lives. Lets say I have a book in me, and I want to get it out. Writing has a process, it takes time, and the time it takes is largely down to human reasons. I can't sit at the computer for a week straight, no bathroom breaks, no food, no sleep, and type out the next great American novel in one go. The cognition fatigues, the body has requirements for sleep and food. Efficiency and speed resent these limitations, encourage us to push past them. I had a professor that described this in a Theological Anthropology class as original sin, denying and pushing past the limitations placed on our nature as finite creatures. She mentioned it mostly to encourage her students to get more sleep, but I find it has myriad uses. LLMs are a product of Silicon Valley, the holy city for modernity. It's high priests proclaim the increase of quality of life that everyone will have by going faster, being more efficient. As a worker, I am compelled to point out that the benefits of speed and efficiency always seem to accrue to the bosses and not so often to the workers, while the physical and mental breakdowns are left for the workers to absorb. I am just not sure exactly *how* an LLM is supposed to make my life *better*. Maybe I'm being obtuse, or dense, but the benefits of a life lived faster and faster are not clear to me, and the benefits of slowing down are increasingly clear.