From TFOriginalScientificArticle: "Additionally, death certificates included a field for reported usual occupation (the occupation in which the decedent spent most of their working life), generally completed by a funeral director with help from the decedent’s informant."

Looking at Figure 1 from the paper, taxi drivers and ambulance drivers are clear lower-bound outliers in terms of risk from Alzheimer's. But they also die a full decade before the eyeballed mean of the population. There are a couple of data points in the middle plot from that figure that have just slightly higher adjusted risk of Alzheimer's, but have lifespans at ca. 80 years that are well *above* the mean. Regrettably, the paper does not discuss those data points -- me, I'd want to have those occupations since they trade a very slightly increased risk of Alzheimer's for 15 more years of life.

And, then, the bit that's also not in the headline and TFS is that taxi drivers and ambulance drivers have *increased* risk of other forms of dementia, especially taxi drivers.

Also, disappointingly, the paper does not show the full ranked list of occupations (even in the supplementary data), highlighting only a small handful (taxi drivers, ambulance drivers, bus drivers, airline pilots, ship captains, and CEOs).

At least someone realized that it would be a good idea to scan outgoing traffic.

One hopes that the internal reaction amongst the leadership was more than just, "bloody hell."

You're making a classic mistake of reversing an association: "this A is a B" does not mean "every B is an A".

Which is to say, you're ignoring the fact that junior associates have attributes and capabilities that LLMs lack, in a wince-inducing oversight. Indulge me with a Sophoclean interrogatory: what can humans do that LLMs cannot?

This is the idea that the anti-AI folks are getting wrong. I'm a vastly more efficient and better writer, of code or of English prose, with an assistant than without. Assistants are, however, imperfect, whether human or artificial.

The current generation of LLM assistants are equivalent in quality of output to having a junior human associate, but at 1/100th the cost. And they're about 100x faster. Is their output perfect? No, but neither is the work-product of a human associate. Does the LLM (or human) output provide me with much of the legwork necessary to create things of clarity, efficiency, and, sometimes, beauty? Oh, hell yes.

I had an argument just last week about didactic presentation with a stubbornly anti-AI professor for whom I have been hired as an instructor for the course he runs. I made the assertion that very shortly, the expectation will be that anyone in that profession who does not use AI will not be bringing their full potential to the task. Personally, I think it is already true that anyone who is fundamentally a communicator of ideas and is not using AI is not achieving their best, but the field does not yet carry the expectation or assumption that everyone has an AI assistant. Not yet, but it will.

Some bamboo species are even halophiles, growing where there is low-quality, brackish water and pulp-yielding trees are rare as hen's teeth.

For services that have no damned business knowing my real birthdate, I always use the start of the Unix epoch: 1/1/1970.

It has two advantages; (1) most calendar picker widgets default to January 1, so it's fewer clicks, and (2), for me at least, it is close enough to being true that no one would ever challenge it, even in person.

The factual reality it ignores is that mistakes by the judicial system in cases covering petty crime, family court, domestic disputes, civil torts, etc., have nothing whatsoever to do with mistakes made in capital cases. Asserting that mistakes made in a local municipal court, which would never see a capital case, is grounds for banning capital punishment is like saying that worms in avocados imported from Mexico means the sale of all fruit grown in the US should be banned.

Conflating capital cases with non-capital cases works against your cause, not for it. If you don't understand that, then you are less likely to achieve your goals. Zealotry -- which is fundamentally a political pose -- is not an efficient means to create change.

That is the most politically saturated comment I've read on Slashdot in years. It completely ignores factual reality in order to push a specific agenda.

Here's a hint: if you want your cause to be taken seriously, don't become blinded by zealotry. In this case, there are plenty enough things wrong with the death penalty that you don't need to make this sort of off-putting overgeneralization, and, by doing so, you are harming your cause, rather than helping it.

The use-case for Concorde on trans-Atlantic passage was cemented for me when my uncle explained that every time he flew from NYC to London to talk to investors about his company, the stock price went up far more than the cost of his trip on the Concorde, and he could be back in time to sleep in his own bed the same day.

Seems like a no-brainer to me. I'm pretty sure (as others have pointed out) that if it weren't for the Continental-caused accident, the Concorde would likely have flown an additional decade.

Looking at current travel, there is enough demand to pack planes on BOS/NYC-SFO, SFO-HND, etc. routes that I'm sure there's enough business money to pay for supersonic service.

Remember kids, when you want to understand at a very basic level what an act does, take the name and turn it to its opposite. The Clarity Act is -- without knowing anything else -- much more likely to be about leagalizing obfuscation than any else.

Distributed power means having 2-3 orders of magnitude more power sources than with centralized systems. That increases the likelihood of an accident by the same factor.

In the US, we have already had near-disaster level nuclear accidents with about 100 total plants. Let's be generous and say that only one was really bad, TMI. That's a 1% failure rate where "failure" means the potential for disaster-level accident. If you want to remind yourself of what disaster-level accidents look like, recall what happened in Ukraine at the Chernobyl power plant, which was caused by human error. And recall that TMI was also caused by human error.

With 2-3 orders of magnitude more nuclear plants at 1% failure rate, that means 100-1000 nuclear disasters in the US, unless, somehow, we are able to engineer plants that are 2-3 orders of magnitude safer, and can find operators that make 2-3 orders of magnitude fewer bone-headed mistakes. As an engineer, albeit a non-nuclear one, I find that a daunting challenge. Yes, we as a society are capable of manufacturing at six and seven nines, but that's when we have lots and lots and lots of practice making things. Right now, nuclear plants only have two nines, with most of the relevant design and construction experience aged out. There aren't enough power plants to be made to develop that expertise, and we'll have plenty of disasters along the way as we learn, where disasters have centuries-long consequences.

So distributed nuclear power? No. Frelling. Thank. You.

Exactly. Just because a design is new with shiny gadgets does not mean it is automatically better than what has been previously on sale.

The sooner the newer generations understand this idea is nothing more than pure marketing hype, the sooner we can break away from and reject the enshittification.

Yours is a far more eloquent way of saying what I had intended to: why is this on Slashdot? Is there any relevance at all? I fail to see it.

If these athletes were coached by AI, well... maybe, but that's a stretch. But they're not; they are just taking more extreme measures to performance enhancement than other athletes. And while I know (and employ) some smart jocks, I had the same experience as you in secondary school, because I, too, was not a jock.

Only recently has it been anything other than relative ranking.