It’s interesting that they use such specific materials. Do you think the material itself makes a big difference to how the characters move, or is most of the animation in the armature?

How much of the benefit here comes from shorter data paths, versus simply being able to fit more memory closer to the accelerator?

That makes the decision seem even stranger. If there really wasn't a qualified Lightning test pilot available, I wonder whether the pressure to get the aircraft moved quickly was the main reason they took the risk. It would be interesting to know what the expected downside of waiting 24–36 hours actually was.

I suspect it was just wanting to get the job done. The difference between this amazing story, and absolutely nothing happening was a fraction of an inch.

I don't know that it was that deep, to be honest.

"Here, you're the electrics guy, go and see if you can reproduce the fault you get when you run that thing up to taxi"

Yeah, the story reads to me less as if it was time pressure and more like they underestimated the risk.

"What could go wrong" etc etc

You've done it yourself.

It's fine, it's a five minute job. Never been a problem before.

Me? No, I've *never* done anything like that, that went wrong in an unexpected way, no no no, not me...

I wonder if this is partly because “pelican riding a bicycle” has become a kind of benchmark prompt by now. If so, could the models actually be getting better at the benchmark rather than getting better at following the prompt?