generated content

#noteinprogress
I don't use AI to write for me. I find that the most boring use of the tool because I'm a writer, I like to write. Writing is how I process my thoughts, and also, depending on what I'm writing about and what my mental state is, it's also how I enter a light trance and transcend being human for a bit.

But I do use AI. Mostly #ChatGPT and sometimes #Claude. I explore other tools as well - #NotebookLM is great for getting an overview of a hyperspecific niche topic to satisfy the ADHD curiosity itch enough that I can get back to work on my current focus projects.

One of my favorite uses of ChatGPT is to come to it with some random idea I had while reading something or watching something or riding my bike or whatever, and using the simulation of conversation to see where that idea might lead and if I want to turn that simulation into something more tangible, that others can read and engage with.

I use the tools critically - I'm fact-checking everything that can be fact-checked and I also regularly ask it to roast me and point out my blindspots - it thinks I'm a bit arrogant and regularly reminds me to consider if I want to dominate or persuade in conversations on social media. I'm still working through that...

One thing that a simulated conversation with a bot is good for is getting reading lists based on those chats. Want to dig into the question, "What does it mean to know when AI is just predicting?" I had ChatGPT generate a couple reading lists for that in the course of a conversation and if you click that link, it'll take you to that note. As I work through the list myself, each title will link to a note with my thoughts on that reading, and probably links to related readings that come up in the course of things.

This is Slow Philosophy. I'm not cramming complex topics into a two-week period one semester - two weeks on "free will" is NOT ENOUGH. But neither is two weeks on any of these topics that I'm interested in exploring.

ChatGPT and other AI tools let me generate content that I can use for self-augmentation aka self-study. Some of that generated content I'll share here - especially the reading lists because it's useful for me to have them on my Obsidian vault to work through them and continue mapping my mind as I continue to augment it with knowledge. Might as well share it as I go. I save my really personal thoughts for a handwritten journal - I can burn what I don't want anyone to ever see, after I've processed in, or perhaps as a rule in my will. So my Obsidian Vaults are the stuff I want to share eventually anyway.

In addition to reading lists, I'll also drop the occasional self-study plan on a topic - ChatGPT often suggests Crash Course stuff, which is GREAT, and gives you the lingo of a field, and then you can build on that going as deep as you want with some bots to point out what direction of thinking leads where.

I'm not outsourcing my thinking, I'm augmenting it. It's our human superpower, the ability to use tools to extend our skills beyond our physical and cognitive limitations. AI is a new way to do that, but we have to be mindful and transparent about how we use it, which is why any generated content in my notes will be tagged with the tool used, and my own writing in notes that have generated content will be tagged/marked as human.