The interface is a bit cumbersome, but it got way better over time and there are some really engaged developers behind it you can always contact. It's intended for biking and hiking, but you can also use it for cars. Meanwhile, they even support android auto.
It's the type of app that grew massively over decades and that offers basically every functionality that you can imagine.
It just hasn't been widely adopted yet. And it might be in each of their particular interests that it continues to stay so for a while. It's basically like p-hacking.
It's impossible to label AI content in general. Actually, I'm quite sick of all the "label AI" mumble. The only thing you can reliably do is to prove that someone is a human.
To say that they're _potentially_ safe by waving at the US Navy is a fallacy for several reasons.
1. It's p-hacking. E. g. with the same technology the Soviets destroyed five of their reactors.
2. The world of civilian operators is completely incomparable.
3. Civilian power plants use different technologies.
> Waste is mostly a solved problem.
Not as far as I know. In Germany, for example, the search for a final disposal site is still completely open-ended, and the first final disposal site will not open until 2074 at the earliest, while, at the same time, the already collapsed storage facilities consume an enormous amount of money. I personally think it is absurd to assume that an underground nuclear waste storage facility can be operated safely over geological time scales.
Needless to say there isn't even a single one worldwide for highly radioactive waste.
And to compare them with coal plants is classical whataboutism. "They can't be bad, because I found something other that's bad as well."
You're right about the minable uranium. That has changed over the last years, so the current estimate is 2080 in a high demand scenario.
But your criticism about the externalized costs falls short as well. Regarding the externalized costs, that is really hard to quantify and I don't know of reliable estimations. How do you want to come up with a number if you don't even know if humans still exist on the planet at that time?
What is clear is that for nuclear energy the majority of the costs is externalized. The bulk of the costs stem from the decommissioning of power plants, final disposal, and accident-related expenses. All three are typically passed on entirely to taxpayers.
The former German vice chancellor even said, he would agree [to build a new nuclear power plant] if <political opponent> found a private operator willing to build a nuclear power plant entirely without government guarantees, subsidies, or liability coverage.
"Model A wins on MMLU. Model B wins on ARC-Challenge. Model C wins on HellaSwag.
At some point you stop trusting any of them—not because benchmarks are meaningless, but because no two of them seem to tell the same story about which model is actually better.
I recommend translating the German version as it is much more detailed.
> "But far more important is the educational value that stems from the spiritual kinship between mathematics and the Third Reich. The fundamental disposition of both is the heroic. […] Both demand service: mathematics demands service to truth, integrity, and precision. […] Both are anti-materialistic. […] Both desire order and discipline; both combat chaos and arbitrariness."
