You've got the elements of the answer right in your comment. Workers in Asia mold plastic with machines at the rates of thousands of pieces an hour. Your hypothetical American putz is making widgets by hand.

Manufacturing costs are no longer primarily driven by labor. They're driven by level of automation, where America has fallen far behind.

We can argue until the cows come home as to the effectiveness of an isolationist stance for bringing manufacturing back to the US. The only thing that will drive on-shoring for certain is a sea change in the way US corporations are managed, by de-incentivising short-term gains. There may be some good ideas in that realm, but I haven't heard them yet.

Those are great hypotheses. Now prove them likely true with preliminary evidence that is rigorously collected so that it might be duplicated by others, and then you get a shot at research money. With that research money, you then need to perform additional data collection and hypothesis testing in a way that, again, is rigorous, using tools and techniques that are widely available, so that others can duplicate and extend your findings. And if the additional data you collect shows that you were wrong, then go back to square one.

Why? Because we need to be certain that the answers are correct before spending massive resources on addressing them. Otherwise we're just pissing in the wind.

Additional hypotheses that have been tossed around here include the massive increase in sedentary behaviors that are well-documented, the reduced vitamin D that comes with lack of sunlight exposure, the lower immunological development that comes from reduced social contact, etc. There are lots and lots of hypotheses. Let's see some data.

Me? My money is on low-grade vitamin D deficiency, weakening the immune system.

It's hard to imagine that their stance will survive legal challenge, but with essentially infinite money, they must have good lawyers who have done the relevant risk analyses.

One aspect of which I can imaging being that once they have trained on copyrighted work, it's effectively impossible to *untrain* the network, so they might be required to pay some modest royalty, but will be able to argue that it is technically / financially infeasible to remove said material.

Well, that sounds very much like, "I can't think of a way it could be otherwise, so it must be true." Here are some counter-examples:

Smoked and dried meat keeps indefinitely.
Salted meat keeps indefinitely.
Water in a slightly porous jug is cooled from evaporation, making it more refreshing.
Snow and ice in a clay pot melts into water when near a fire.
Clay pots are excellent for carrying things from point A to point B, no matter what they are.

And that's with 30 seconds of thinking. Thus, I assert that your assumptions are incorrect.

Ah, my favorite among such is the warning from a hair drier I bought some years ago. It said: "do not use while sleeping."

That's an interesting link, thank you.
The "nuances or causal factors" do show up -- somewhat -- farther down that report:

I can't tell if you're being intentionally dense or not, so I'll err on the side of naivete. The construction and operational validation has been paid for, which is the largest part of the cost. The ongoing costs are things like salaries, materials and supplies for subsequent operation, maintenance, and improvements, which are far smaller.

There is no scientist I have ever met who thought LIGO was, in the end, a poor choice of investment of national research funds. There were plenty prior to its stunning first detection (myself included) who thought they were chasing ghosts, but all of those doubters have been converted. The important thing to understand is that LIGO's contributions weren't just detection of a black hole merger (in itself, a hugely important event because it demonstrated the hypothesized existence of gravitational waves), but the establishment of a new field of astronomy based on gravitation, an entirely new means to observe the universe that provides information previously completely unobtainable. Our eyes have been opened where we were previously blind, and the ongoing results are, and continue to be, astounding.

There's a nice fact sheet summary at: https://ancillary-proxy.atarimworker.io?url=https%3A%2F%2Fwww.ligo.caltech.edu%2Fs...

I've been using it to write grant applications, and I share your opinion. It frequently makes mistakes (and 5 is worse in many ways than 4o). While it can certainly be used to create a rough draft of a document, the result is similar to what you would expect from a junior associate, with the same kinds of mistakes that create an, "OMG, no," response in the reader when it starts to make things up.

There was a lot of talk about how rapidly it would accelerate in performance. That progress seems to have stalled this year. I have a hard time thinking that we've started to see the ultimate asymptote of performance, but it seems like we've hit a region where the easy, early gains have all been made.

I think you are thinking of geo stationary sats. I worked with a sat HD TV link bandwitch was about 6 Megabits per second and it was 400 ms plus. Kuiper like startlink are in LEO. So times are less than 100mS. YMMV

EBay is your friend. Plenty of them available there!

Fair. MY point is that I don't trust that source because whomever/whatever wrote it made a logical error.

OK. But copied the quote from where? I can't find a source worded like that.

How is 868 a decrease from 813?

It isn't just that new factories are fully automated, it's that they must be so in order to be competitive. The only way manufacturing in America can be competitive to manufacturing in low-wage / already-automated countries is through automation. The era of factories being built with lots of human employees is gone, and even in the industries we thought would never fall to automation, robotics is making significant inroads. Why? People are just too expensive.

There is a lot of backlash against p-values. I won't enter that argument here.

But, accepted standards are that p-values need to be below 0.05 in order to be taken seriously in a biological study, and, even then, they aren't that strong until they get substantially smaller. (Note that in physics, they've been burned so many times by crappy significance values that the standard there is 5 sigma, or about p = 0.0000003.)

This paper, according to the summary, reported an effect at p = 0.079. That's substantially above p = 0.05. Me, I probably wouldn't publish anything at that p-value, but take my licks and move on to the next study, having learned how to design better experiments.