They want to "defend against memory safety vulnerabilities?" I assume that they're talking about buffer overflows, if nothing else, and I can think of a couple of ways to prevent them: 1) non-von Neumann architecture; or, and here I'm going really crazy, I know, with an idea that'd disrupt the entire industry: 2) stop using bloody C.

"Stable equilibrum temperature" implies that the heat from the fuel, including any decay heat, is going somewhere. If that (presumably passive) process can continue indefinitely without risk to the outside world, and if the equilibrium temperature is below any temperature at which melting of the plumbing or new and exciting chemical reactions could occur, then the design would appear not to be vulnerable to a loss of power to the pumps.

Heating up the coolant naturally slows down the reaction.

Except that a runaway reaction isn't the problem here. The reactor reportedly scrammed as designed. As at Three Mile Island, all of this trouble is being caused by the decay heat alone.

I note that the Wikipedia illustration of a molten salt reactor design shows a pump in the cooling system, so it's not immediately clear to me that a molten salt reactor would survive the loss of emergency power reported to have occurred here. And even in the case of a design where the cooling is driven entirely by convection — with, what, giant air-cooled heat-exchangers? — there's the question of earthquake damage to the plumbing, which I suppose is the other obvious possible cause of the current problem.

...that this article is baseless fantasy. Half of it's gibberish: what does "cities heated by servers," even mean? The other half ignores what's known to be possible, with the holographic projections popping out of phones within four years being the most obvious clanger. How's that supposed to work? Like in Star Wars, of course, which is to say only as a special effect in a movie.

So no XML...

I'm not really a COBOL-lover, but you're starting to sell me on it.

"I don't get paid for work I did two decades ago. Why should you?"

A possible response: "Because work that I did two decades ago is still valued and in demand, while nobody cares what you did yesterday."

We use all kinds of contextual queues...

Hah! Well done.

When one of the questions is "Are you empathetic?" and the answer "yes" results in your being scored as empathetic, the test is, as others have noted, unlikely to provide any insight. The only way this little test works is as a sort of meta-test: if you can't pick a result and get it on the first try, you're not very good at imagining what the person who designed it was thinking.

Just by answering each question by giving the strongest response in what I judged to be the appropriate direction, I was able to score 70/70 on the empathy scale on the very first try. For my second trick, I successfully scored the minimum possible, an angry red 1/5 on each question. I didn't even bother to systematically check my previous friendly green 5/5 answers and reverse them.

For what it's worth, I then made a half-way honest attempt, without any real soul-searching, to pick responses that I felt described me fairly, picking the middle of the scale on the most egregiously ambiguous statements, and I scored bang in the middle: 51/70. I think it's safe to say that the results mean nothing, alas, so I still can't settle the question of whether I'm an android or not.

This particular catastrophe is nothing on the geological time scale.

What a pointless truism that is. The worst atrocities in human history, combined, are "nothing on the geological time scale."

It would have been correct if he had said "many more datum", but that's just weird.

No, that would be like saying "many more apple" — "datum" is emphatically singular; it's "data" that people like to play fast and loose with.

The opposite case of "dice" might be of interest to ESL learners: the singular form is "die," but many people say "a dice." Are they ignorant of their mistake or are they speaking a newer dialect of English? Probably.