joha4270

Real projects has a way of encountering unexpected troubles in a way thay paper ones dont.

On top of that there is the energy security problem, even if Morocco is less likey than Russia to try and play tricks.

I wish them all luck, but we can't forget the advantages projects has before leaving Excel.

> By the way, do you know what cost the most money on the HPC project? Loan interests, by a very large margin.

That sounds very interesting, do you happen to have a source nearby? I would love to have that one in my back pocket next time i end up in a discussion on nuclear power.

And so the "AI-canaries" are starting to appear. And I'm sure somebody will soon try to figure out how to avoid them automatically.

I'm sure it will be a fascinating cat and mouse game, except for the part that communication's S/N ratio will get even lower...

Could you expand on what

> conversationally fluent in the theory behind writing code

means?

It might be my insufficient command of the English language, or I might be outing myself as being outside said group, but I'm unsure what that means. Is this just referring to a vocabulary for discussing the structure and creation of software, or is there a deeper mystery I have not yet grasped?

I don't really think that's a fair assessment, since most of it is in the core. That's some very long and heat resistant roots.

I also can't help but wonder, could trees even use iron if it was plentiful in the upper crust? You need a lot of energy to separate iron oxide into elemental iron. Betting against what evolution can make is usually a bad idea, but that would be a neat trick.

Don't be ridiculous. Yeah, we truly need consequences for corporations (and arguably their leaders) that isn't just cost of doing business, but the upper end of your suggested sentence is a span longer than written history.

Wanting a harsh sentence is fine, but lets limit it something we can likely see done before the end of our civilization in its current form.

I don't think Dragon has spare seats[1][2]. So sure, if one of the Starliner astronauts develops a medical emergency, they might be able to take Dragon on a ride down. But that leaves two other astronauts that has to ride down Starliner.

That said, yes I agree that the headline is slanted somewhat more anti-boeing than the public knowledge supports. Yes, the delay has been long and not a great look, but double and triple checking an experimental spacecraft with some _probably small_ (taking the article as gospel here) is a very sound course of action if you don't have somewhere you need to be.

[1]: https://en.wikipedia.org/wiki/Crew_Dragon_Endeavour (4 passengers this flight) [2]: https://en.wikipedia.org/wiki/SpaceX_Dragon_2#Crew_Dragon (4 seats)

We do call current intel CPU x86.

As so many things, the organization we now call EU has changed over time. Pretending it formed ex nihilo in 1993 is at least as silly as ignoring what came before.

Reality is messy and drawing hard lines can easily discard a lot of nuance. But if I have to, I would place the foundation of the EU in the post-war era, not at the Maastricht Treaty

The European Coal and Steel Community, which in many ways is the precursor organization to EU was formed in 1951, with the explicit goal of linking the German and French economies to such a degree that they would be incapable of waging war against each other.

It sounds a little high, but not massively so.

I couldn't find fuel weight in a BWR-3, but a RBMK reactor (Chernobyl) contains between 150 and 200 tons of Uranium. Now, a RBMK runs natural uranium, so you're going to be using more fuel for the same power, but it suggests the number isn't totally whack.

I would not be surprised by it being a typo, but depending on what you include in "melted fuel" (fuel assembly, reactor parts, etc), 880 tons isn't outrageously high either.

Phone networks can already do this. And its true broadcast, you can decode it passively without ever connecting to the network, so no SIM/transmission required.

I have no idea this is actually enabled in networks in practice, but the specification exists.

I would be a little sceptical about that figure. 3 million dollars per second is around the world GDP.

I get it, AI training is expensive, but I don't believe it's that expensive

This article looks like an ad too me.

The article and the linked original source[1] is somewhat sparse on technical details, but if I divine it correctly, what they're making is essentially a GPU, with its compute units being RISC-V cores instead of some proprietary ISA.

I predict this will be about as successful, as when Intel tried it a decade ago with the Xeon Phi. But lets see

[1]: https://www.x-silicon.com/news/