Look at the second paragraph under "How to Apply These Terms to Your New Programs". That section is very different, with numbered spacers inserted into it. I'm sure that's some sort of joke or reference, though.
This article doesn't describe it in detail. One scenario imaginable would be that they ran their model trained on non-full moon data for an evaluation on a full moon day. Which means the model would simply apply it's learned "optimal" action policy in a different environment, where the previously learned action policy doesn't lead to good scores anymore.
I have never seen the big main lights using blue instead of green anywhere in Japan; the traffic lights for pedestrians however were blue in some places.
People on this platform, I assume, are familiar with the scientific consensus on these topics and generally side with the science. But one has to understand, that these decisions are not made directly by scientists but by politicians who have scientific advisors. And as politicians, they will factor in a lot more than just the hard science on a topic, which can result in seemingly unscientific decisions that go against the "obvious" findings.
It confuses me that you used a video with that much high-pitched background noise in the audio to showcase your tool - in the end, both version seem a bit low-quality because of the noisy audio.
I've been using Typer for my projects recently and I have to agree, it's great! My only gripe with it is that its autocompletion-installation arguments can't be turned off, so it "pollutes" your help text in a way.
It's not a big deal, but especially for smaller tools with small CLIs this annoys me a bit.