Everyone has an opinion on generative AI (AI) and I want to clarify my position in the interest of transparency.
To be upfront, nothing on my website, from the coding of the site to the essays I write, is AI-generated. It is all made by me or sourced by an actual human. However, I am not AI-hostile. If AI can speed up certain processes in my writing, such as comparing data, I find no reason to ignore a tool at my disposal.
LET ME REITERATE. NONE of the ACTUAL WRITING on this website is AI-generated. Tedious tasks like comparing data MAY be done with AI-generated tools, but the actual collection of that data will always be by hand because I don't trust AI with that task. My data has always been open-source, and I welcome anyone with actual programming skill to double-check my analysis. I am fully aware that despite my rigorous testing, AI-generated programs always have a margin of error inherent in them because they have not been checked by a skilled human.
AI art and writing is soulless and will never be superior to human-made art. I have written an essay on The Emoji Movie - raw quality =/= meaning, importance, and enjoyment. I analysed The Emoji Movie because I wanted to connect with the humans behind it, the meaning they intended for it and how that meaning was communicated. You can't do this with AI. There's no point in analysing the deeper meaning of, say, an AI-generated sentence, because the AI doesn't know what deeper meaning is. It is generating based on patterns in its dataset - mathematical but meaningless. There is no human thought, and thus, no reason to analyse it outside of the data it outputs.
This is on top of ethical issues. LLMs are trained on stolen data and guzzle energy like a motherfucker. This is just exploitative. Exploitative of the people whose work is stolen, exploitative of the planet. I don't believe either are inherent issues in the technology itself. There are attempts to mitigate the environmental impact, and given input that is public domain and/or made of wholly consenting parties, there are no ethical issues to AI. I don't think there are AIs like this in existence, however. The fact that people and corporations steal people's work and profit off of it is bad!
However, I don't live in an AI-free world. I will use AI as a tool if it can help me with my hobbies (my hobby is writing, not data processing!), but I will also use AI for non-profit fun. I liken this relationship to the relationship the Mayans had with wheels. They had wheels, you know - on toys. What use do wheels have in a society living on a mountain, where the terrain is too rocky and too steep for wheels to be useful? I'm an artist and a writer. I don't need AI to do this for me. I find enjoyment in the process as much as the finished product. I don't need a wheel on my mountain.
What is this "fun" I speak of, then? AI RP. To be clear, I don't use AI chatbots to replace real human connection, nor do I use it to write anything that I publish here, beyond what is in the bots I publish on sites like Xoul and WyvernChat. Its a fun time-waster and somewhat intellectually challenging due to having to figure out the best way to phrase input for the LLM to produce the output I want. I don't give money to these services and the code for my bots are public. My lorebooks on WyvernChat are private less to protect them and more because I'm self-conscious about flooding the site with them. If you want to see them, ask! I say this because, to paraphrase someone on the WyvernChat Discord, the LLMs are trained on stolen data, so it's weird to be possessive over what you feed into it.
But this is the typical extent of my use of AI. Yes, I've used ChatGPT to make a program for me before, but I did that after asking skilled human programmers around me to do it for me (and I understand why they didn't, I am broke, why do you think I'm still on a .neocities.org address). People often say that the only thing they want AI in their life for is to make it easier to live their lives - not to do their hobbies for them. I am a writer, an artist, an an academic. If AI can sort through all of my manually-inputted data for me and cut weeks off the time spent processing that data, all so I can get to the actual analysis and writing of what that data means quicker? I think that's a good use for it.
In solidarity with anti-AI writers, however, I do not consent to my writings being used for training, but I also understand it's probably already been done. Making things disability-accessible also makes them AI-accessible. I know this as a risk and have accepted it.
So yeah. This site will never be written with AI because it would defeat the purpose of the site. If I wanted to push out slop, I'd at least try to make money off of it, and I'd at LEAST try to publish it on a site that more than five people would even see, and I'd publish it more frequently than - what, one month at minimum?