The default markers have the distinction of being plentiful AND consistently ineffective at representing ANY of the conditions I encounter in any of the systems I run! 😄 (Which is a lot of them.)

I am a fan of and I use those JinxShadow tokens, but I'd love if Roll20 updated the defaults!

Lawful Good

You are, in part, a science communicator, and science is typically methodical and method is lawful.

You are also a community builder. The "the internet is a place we live and I want it to be good, useful, and safe" worldview is a very lawful position.

You probably encounter more AI writing because of the places you most often visit online vs. the others mediums and caught on to the patterns quick, especially because you REALLY know what human writing, and human menswear content, looks like.

With more exposure, recognizing AI images/video is easy

Even absent of accuracy concerns, I only value communication as a holistic product of a person's mind, warts and all. I find that a person's ability to articulate something, and any lumps along the way, is fundamental to the interaction. I don't want the polished-up version spit out by a machine.

Use of LLMs can lead to something like the Gell-Mann amnesia effect, i.e. we tend to be less critical of text on topics we are uninformed about despite often seeing inaccuracies in content related to topics we're familiar with. LLMs replace context you can use to evaluate legitimacy with novelty.

If the goal is to win more often, you can't beat practice. While practicing, I find the secret is to work backwards. Try to find out what actions or circumstances the winner usually finds themselves in and figure out the path to get there.