AI Disclosure

My role is, and has been, to provide my experience and perspective on AI to people on the Internet. That’s what my channel became known for, and what my analytics tell me people still come to my channel for. In order to keep up to date on how well (or not) AI is working, I have used it and will continue to use it. If you’re someone who believes that no one (or no creator) should ever use AI, I’m not the creator for you. I’m very specific about what I will and won’t do with it, though. Hence this disclosure.

How I will (or won’t) use AI in my content creation (none of these are sponsored):

I use AI to create closed captions for my videos (although I always have to edit them, because they’re never right the first time). I also use AI to level the audio for my videos and to remove extraneous sounds like breathing, air conditioning, etc.

I routinely use AI to search for things on the Internet (typically perplexity), but I click through to the sources, and don’t rely on what the AI says (because the answers from the AI aren’t reflected in the sources, more often than not).

I use AI like a thesaurus - when I can’t think of the right word I want, or if I have a phrase that sounds clunky and I want to make it flow better (though it doesn’t always help). I never let it write scripts for me, though. (Full disclosure - I’ve actually had it try to generate parts of scripts for me, and they’re horrible).

When I have a lot of sources to go through as sources for a video, I will upload them to NotebookLM to find and pull quotes for me, but I always double check the quotes (and generally I put screenshots of the quotes in the videos themselves).

I use Google Gemini to suggest video metadata like descriptions, titles, keywords, hashtags and chapter markers.

I will occasionally use AI (lately nanobanana) to make thumbnails. I wouldn’t use AI to avoid paying an artist, though. It’s either AI or my crappy photoshop skills. If I ever get to the point where I start paying professionals to help create my content (editors, thumbnail creators, etc), then I’ll stop using AI for that.

And, of course, any time I’m evaluating or reviewing an AI, I’m generating the content (often code) that I’m reviewing (I expect that goes without saying, but I’m trying to be transparent).

That’s all I can think of. I’ll update this page if I think of another way I use AI, or if I add any new AI workflows.