Some feedback:

I could not catch that the circle meant the sun. I did catch seconds, minutes, day, month, and I deduced TK meant Tokyo and SF meant San Francisco and could not deduce the city ZG represents. So, I don' think you can actually be free of baggage. Some minor things bugging me: - The lines at the bottom don't align! Really annoying, I spent a lot of time trying to understand it meant something or not. - Dark mode shows only 1 city. I could not understand what TK meant until I saw SF.

Some advice:

- Change the pictoric representation of the sun for something a bit more "sunny". Some flames feel like something not cultural (since there literally is solar activity at the surface), so aliens could get that it is the sun. - In the intro you say there is a puzzle, but you don't say what the puzzle is. I thought it might be some interactive thing. It was only until I gave up I start reading the explanation and got that the puzzle is to understand how the clock works, given it has no cultural baggage. Then I went back to try to understand it. Maybe add a little explanation somewhere saying the intention.

Why would people want a standard to be controlled by a private company? I don't think the "Open-Sourcing" of it says enough. How does licensing work with formats or standards?

But in this case, it is really hard to know if a model is identifying "correct answers" reliably. A lot of answers are really hard to qualify as correct or not when written by humans, much more when written by a machine trying to trick readers into thinking the answer is correct. It can be done, but I doubt LLM are being trained to identify the subtle differences between those types of potential answers.

Not caring at all about false positives, which by the way are very common, enters the category of incompetence for me.

I don't think you need to improve that much current LLM so they can detect actual harm threats or hate speech from any other type of communication. And I think those should be the only sort of banned speech.

And if facebook wants to impose additional censorship rules, then it should at least clearly list them, and make the moderator AI explain what are the violated rules, and give the possibility to appeal in case it is doing wrong.

Any other type of bot moderation should be unacceptable.

But this implies that people at facebook believe so much in their AI that there is no way at all to appeal what it does to a human eventually. Not even for doing learning reinforcement they have human people to review eventually some post that a person keep saying the AI is flagging incorrectly.

Either they trust too much in the AI or they are incompetent.