With regard to research/data, you're fact-checking the robot, right? They often say untrue things because they're not capable of knowing whether or not something is true.

Check out goblin.tools, it's made exactly for that and can do much more really useful stuff.

Oh, I would break down my tickets into tiny baby steps, to the point it's so utterly ridiculous from the outside looking in that my colleagues and manager would comment on my "work journal" and are concerned about the "way overcommitted wasted effort."

It of course, helps a whole lot despite the amount of additional added hours I would put into that. Each little checkbox is a much needed dopamine trickle to keep me on task, and the whole thing seems far less daunting once I've been able to sit down and break things down into atomic subtasks—checkpoints—that I could stop at and still be able to come back later on without losing track of where I'm at mid-ticket. And not ending up debugging myself round and round in circles as I forget what I already did before.

I didn't use to do that though, nor did most of my systems even exist in any tangible form really. They were the products of my brainstorming and research on how to get around my ADHD and not things that I would've naturally settled on on my own. So, if I'm understanding what (this aspect of) autism is correctly. I'm only ADHD as well, at least on this (and many other so far) aspect.

Edit: reading comprehension failure. Um, I haven't been using AI for the breaking down of. Maybe it's something I could try?