I recall an explanation of charge separation in clouds when waterdrops would form a thin layer of ice which would shatter in tiny pieces. This would leave a charged waterdrop behind. The ice fragments and the waterdrop would move at very different speeds and this would lead to separation of charge.

I wonder if those water drops have the right charge for corrosion.

Don't give them ideas!

it is about how you read the data.If you make things safer people make the calculation they can relax a bit more and be distracted. if you have lane assist you can pay less attention to your lane and a bit more to your smartphone.

In a way a lot of the modern safety gadgets are 'multitasking assistance' that allow you to think about other things.

it is possible to calculate the total amount of radiation she was exposed to, and to map that to tables that link radiation exposure to risk of cancer and death.

The tables can probably be challenged - I recall indications that the risks of long term low intensity exposure were overrated. But regardless of that, there will be a question of 'how high does the number for risk of getting cancer have to be before the judge decides the state has to pay compensation?

At least my AI girlfriend has real boobs,

Why are all the posters making this comment downvoted as troll? The part about the roundups is getting ahead of itself but I think the part where everyone is made permanently identifiable on the web in order to control them should be central t the discussion

but it isn't.

I suppose they want some combination of acquiring a broad skillset and showing it off. Just these robot games already have dozens of competitions going on.
And do they want robots that can run fast? I think so . As a kid I already imagined all terrain walking machines that could move real fast.

It is always propaganda but usually inspired by real events. In this case even that is not sure.

I only came because it'd been a long time.

Pretty sure they will be able to do it in 6s soon, about what the fastest mammals can do.

At 4.5 seconds you're accelerating at 1G on average, and you're running so there would be an air phase. So the design will have to start paying special care to grip.

But parent post's remark would be statistically relevant.

They deal with it by adding 'likely' in the conclusions.

I believe their conclusions though. But they also say 'significant signs of AI authorship' which is very stretchable and more cautious compared to other studies.

I always was fond of the joke about the solar probe that only flew at night to protect it against the sun.