@OwenReynolds I don't think those posts on Twitter are notable claims by our rules. But the original claim by the NY Times that these are real X-Rays is clearly notable, so it doesn't matter. It does not seem entirely unreasonable to doubt these images, some parts look a bit dubious to me. Doesn't mean they're fake, I don't know anything about interpreting X-rays and what looks like it might be manipulation might be simply a consequence of how this method works.

The main issue here is whether these cells are from a cell line, which is very likely and cell lines have been the source of similar arguments before. I couldn't find a reliable reference for this case in a quick search, so I didn't write an answer. The difference with a cell line is that those are reproduced essentially unlimited in the lab so there would not be foreskin in the cream but cells derived from a single foreskin harvested years or decades ago.

Jail breaking LLMs is something entirely different than "using sophisticated AI tools" for hacking. I think the quote here is very misleading, but we should probably focus on the second paragraph alone here. The first one is not actually related at all, it's just bad journalism. Using AI tools for hacking does not necessarily imply that LLMs are a part of either the attacker or the target (though I suspect the implication here is that genAI hacking tools were used, which I personally strongly doubt).

@quarague your assertion that the main explosive power comes from the fusion part is not cited and I'm reasonably sure it is not true for many modern nuclear weapons (though there are certainly some where it is true).

Modern thermonuclear bombs have much larger yields than the original fission bombs, you cannot ignore that aspect. You'd also have to distinguish whether you're talking about fission-fusion-fission bombs with a uranium tamper or bombs with an inert tamper. And then you'd have to show which types are actually most common in modern nuclear arsenals to answer this question.

Can you please provide a source for the 37% claim, I'm not sure that is notable. The specific arrest quoted there is notable, but that is not your actual question.

Are you only interested in whether it is possible in general to do this, or also in the specific efficiency claims? The former is very easy to answer, the latter might be more difficult

The general impression I've gotten from military and political experts is that they consider any statements like this from participants in a war suspect without independent confirmation, even if they otherwise trust the government or institution. Simply because there are huge incentives to publish information that supports your own side. An example would be the war against Ukraine where while experts might consider Russian statements entirely fiction, they also don't take Ukrainian ones at face value.

That article states that the "Unchained At Last" data was compiled from marriage certificate data for 38 states, which seems to me the most robust data possible. The remaining states are extrapolations, but that should not change anything fundamental here. I'm not sure what kind of data you expect here

@CDR I removed it. Pretty much any question related to this conflict is something I consider problematic for the HNQ. Many of the topics are disturbing and not appropriate for the HNQ, and the massive attention makes moderation a lot harder. These question also tend to attract a lot of problematic answers and comments from the HNQ if they are left there.

@DavePhD I had to ask a CM to redact the image, that likely caused collateral damage. And all that wouldn't have been necessary if you hadn't posted that image in the first place.

Do not post any images of these events in answers here. Images of dead people and other gore are not acceptable. If they are absolutely necessary for the answer you can put them behind a non-clickable link.