10 Watts times 60 minutes is 10 Watt hours = 10 Wh (or 600 Watt minutes).
In the parent post all results are off by a factor of 60. The conclusions are mostly correct.

As noted or hinted upon by others, the concepts challenged in this paper (i.e. general relativity) are very well established and tested experimentally, e.g., in the context of GPS. Conversely, the question of what negative masses or mass densities would imply is obvious, so that probably every single physics student or professor has thought about it. Still, nobody seems to have found any traces of negative masses.

Thus, the authors suggestions should be viewed as (potentially) somewhat interesting speculation or an empirical approach for fitting cosmological data.

I dont think there is anything new that the general public could/should learn from this news or the article.

The exact same gesture could be activated as an accessibility feature for quite some time even on older models.

The implementation of the powers of ChaptGPT and assoicated technologies into the Microsoft office suite will be a complete game changer. Nobody (in his right mind) will manually organize emails when an AI can perfectly sort emails, alert you of important infos and deadlines and suggest answers or actions.

This has no more similarity to Clippy than a Tesla has to a toy car.

He is a teen (19) - the latest cheating evidence is from 2020.

NVIDIA A100 cards do not have video outputs (like HDMI) and, I think, also no raytracing units. So they are definitely not video cards and not really (or mostly) GPUs.

Click on "expand" in the statistics. You quoted the average of the Euro countries (excluding, e.g., the UK). Germany, by the way, has 6.4% (according to this chart).

How can such a bug in a security sensitive component of OS-X be overlooked in testing?

There's been consideration of self-mutating algorithms where the hash of the previous block determines the problem to be solved for the next, functionally making CPU/GPU hardware the most efficient way of solving it (because you have to be generic). That's a damned difficult research question, though, not something we have a ready to go answer for.

Although seriously, it's generally considered in poor taste to award Darwins for a person or group who take out innocent bystanders along with themselves. After all, they're not necessarily improving the gene pool if they take others out too.

"It is a damn poor mind indeed which can't think of at least two ways to spell any word."
--Andrew Jackson

Don't forget Antares was missing from the night sky; I cling to my theory that it going supernova damaged to ozone layer sufficiently to precipitate an ice-age that dropped the ocean levels, closing Gibraltar.

I don't know, an EE PhD might...

As a (candidate for) PhD CS, though, I'm fairly certain if I turned up with a screwdriver one day, people would panic and/or probably tackle me to the ground.

Well, if they're doing this properly, it shouldn't be about whether the student learnt the material, but how.

It should be used to show:

Students who aren't engaging with the material, and may require early intervention
Levels of interest in the material (would different material suit the learners better?)
Problems with the material (are there particular parts many learners highlight and/or comment on? Could indicate confusion, for example)

Why is it disconcerting?

I mean... yes, it can be mis-used. The data should be used to flag up pupils who may be struggling, but will also flag those who may already know the material, but just because data could be incorrectly used doesn't make it inherently worrying.

Does it?