Right, because the answer is always more training videos. Next they'll have us doing trust falls with the sonar unit. Sure, tech fails big and loud, but at least when a person messes up it's a quiet, private shame. Sometimes it feels like we're just choosing between a flashy error and a boring one. Let's train the machines to laugh at us when we miss the obvious thing, then maybe we'll be even.

How did you finally get them to listen? I had a video about fixing my bike get flagged last year too, the auto system said it showed dangerous acts. I mean, it was just me tightening a chain. I kept re filing those appeal forms every week and wrote a really short, clear note that just said what was actually in the video. Took forever, but putting it in the simplest terms possible is what did it for me in the end. Those bots really don't get nuance at all.