>> or liquid nitrogen, or hydrogen,

> Not some stupid liquid hydrogen

Eeep! I'd intended the "liquid" to be scoped by the comma. :) Though I'd expect liquid hydrogen to be even more popular than the pop of hydrogen gas. The play doh freezes, and breaks like glass, and then... people like big explosions.

Neat Maus show - thanks.

But... consider the difference in testing culture between app development, and education video development. I tried creating some education video, using software style iterative development, and street guerilla usability testing. Which turned up all sorts of unexpected failure modes (say "millimeters", and some have traumatic flashbacks to learning metric in school; say "fun story about viruses!", and for some, that has the emotional loading of "fun story about genocide!"). Later, I was chatting about it with someone from WGBH, Boston's PBS station. Which makes a lot of education videos. Their comment was something like, yes, we'd love to try iterative development and testing, and we will... just as soon as we find anyone willing to fund it. So an underperformance of science education videos seems no more surprising than an underperformance of large-scale waterfall software projects.

> kids are incredibly primed to recognise and cherish brands

So I'd rephrase this as: the toy industry devotes lots of effort and money and testing, to creating things kids will recognize and cherish... and the science education video industry simply doesn't.

As for the "silly" videos, imagine you were playing with kids. You might do something much like in those "silly" videos - hands on, rich in phenomena (play doh tearing, describing objects, making choices, and on and on)... fun and interesting. You wouldn't do "Now I will show you a still photograph of part of a bottle of sun screen. Note the number 10..." blah blah blah (example from the Mouse TV 1 video). I myself found the silly videos more engaging, despite an ignorance of, and distaste for, the branding and characters. So I suggest at least some of the appeal is elsewhere.

I greatly enjoy going to research talks. I generally don't bother watching science education videos. There's a richness in one that's missing from the other. And consider molecular biology thesis defenses (went to one yesterday), which at least around MIT and Harvard, end with a long thank you section. Something I've not seen in other fields. It extends to family, and significant others, and pets, and friends. It is illustrated with funny photos, and is leavened with inside jokes, and hints of their challenges and character. A moment of standing at this major milestone in their life, and looking back. It is just a richly textured few minutes. Not all defense acknowledgements are that notable. And yes, there are some science education videos which try to capture a bit of that in interviews. But you don't have people crying, or slips like "thank you to my husband <name of their professor instead>" as the room cracks up. It could be valuable educationally, to show those segments to kids. Especially with students from underrepresented groups, or no-college families. So they could see, that person is like me - that could be me in a few years, and it looks like they had fun getting there. But like so much else in science research, which might be of value in science education, the incentives along the pipeline to get them from one to the other, are absent or dysfunctional.

> this flood of trash

So perhaps a more upbeat perspective, might be that society is investing lots of creativity and resources, to discovering how to create compelling media and storytelling. Yay! And it's now up to people who care about science education, to create content which leverages those insights.

Hmm, I wonder how VR/AR will impact this silly video niche... hands-on direct manipulation play-doh egg surprises?