Most desktop 3D printers are Chinese. The American ones (made in China) mostly went away because they cost 10x as much, were near impossible to repair, and some were drowning in DRM.

If American companies don't stop phoning it in and demanding huge margins (or running with terrible efficiency), they will be left behind.

War doesn't always start with a clear-headed, cold-blooded weighing of national interests. In fact I'd say that's the more the exception than the rule. Historically it's quite common for a country to start a war that in retrospect looks stupid from the standpoint of national interests.

Of course peaceful initiatives can be just as badly thought ought. We quite *deliberately* chose to tie our economy to China; I remember this quite distinctly. Although nobody anticipated the speed or completeness of the interdependency that would folow, everybody understood that we were choosing to head that way. The argument was a purely ideological one, whether interdependency per se was a *good* thing. And, as far as it goes, the argument was sound. If you don't nitpick too much, it worked out just as planned.

The thing that we really didn't put much thought into was *who it was we were choosing to become interdependent with*. China is, not to put too fine a point on it, an unstable and very dangerous powder keg. There is no rule of law; laws are enforced selectively by officials tied to an unaccountable and unrestrainable political party. There is no freedom of information, which means among other things you don't get economic data you can trust. The system is prone to sudden, opaque power shifts and the emergence of strong men who are legally, and sometimes politically unrestrained with respect to policy and military affairs.

And now we'd really like a little more distance from that powder keg, but our interdependence is the main thing that's stabilizing the situation. At least in the short term, until somebody does something that, in restrospect, will look really stupid. Which is inevitable, eventually.

Presumably critters evolved to deal with noises that naturally and regularly occur in their native habitats.

This doesn't mean that natural noises that aren't regularly part of their normal habitat can't harm them. It's possible that animals whose range naturally overlaps the periodical cicadas do get harmed by that noise, but the harm is not significant enough to exert selective evolutionary pressure.

So natural isn't necessarily benign. Nor, do I think, is *unnatural* necessarily harmful. But dose does makes the poison, and cars do make a *lot* of noise. It's pretty well established that humans overexposed to car noise can develop health problems like cardiovascular disease. Since CVD mainly kills and disables people after their reproductive years, don't expect populations to evolve a biological tolerance for car noise though.

He's moving some assets into US companies because they're innovative. Fair enough.

He thinks they're innovative because they've got more hustle. OK. That's almost circular.

He thinks they've got more hustle because Americans work longer hours. That doesn't follow at all.

Sometimes you work longer hours because the boss forces you to, and you are giving him as little for the time as possible. Sometimes you work longer hours because you're disorganized, bad at planning and managing your time. I've seen that often enough. If hours worked equals hustle equal innovation, he should be putting his money into Cambodia, where workers put in 40% more hours per year than Americans. Sweden and Switzerland rank higher than the US in the Global Innovation Index, even though people in those countries work a *lot* less.

Innovation for a country is multifactorial. Wealth and education matter. Attractiveness to foreign investment; rule of law; those are really important things where America excels. Even sheer size makes a difference; being part of a massive integrated market is a huge boost to both the US and the EU. Sure, work ethic matters, but work *hours* is a lousy proxy for that. In some countries people put in six hours of honest hard toil each day then go home. Do they have less work ethic than a country where people spend ten hours a day at work but much of that "lying flat"?

They're laying out their plans for up to the end of the decade. You think this is just a fraction of their NA plans, that they're hiding things from their giant "Look, We're Doing EVs!" press release?

If it works for them, are they wrong?

Except as labor standards drop, your choice is another job that does the same thing. About 17% of American workers don't have fixed hours or guaranteed workdays, which makes planning for work/life balance a farce, and the old standby of getting a second job to make ends meet is impossible.

73% of young Americans live paycheck to paycheck, 20% of whom have no savings at all and many of them have to spend 50% of their income on housing. This means they don't really have the ability to quit their job and look for another job where working conditions exceed the minimum legally allowable standards. Which is why legally enforced minimum standards are important. We need those young people to step up and start making babies.

Fertility rates have dropped in the US from roughly replacement (2.1 children/woman) to a catstrophically low 1.6. The US population would already be contracting were it not for immigration. Now a lot of this is social changes -- women choosing to delay childbearing to start a career. But consider South Korea, which has the lowest fertility rate in the world at 0.8. They're a much more conservative society than we are so it's not changes in attitudes that's driving that. The reason their fertility rate is so low is that they take people in their prime childbearing years and work them like dogs, in return for little prospect of economic security.

Don't you think if those young Koreans would quit their job and choose a higher paying job that gives them more leisure time if they could?

When I started working in the 1980s, getting your first job was like stepping onto an escalator that would carry you up to higher economic status. It's not like that now for the youngest generation of workers; it's more like stepping onto a treadmill. When we start to look to that generation to replenish the US population, our fertility rate is going to sink like a rock. The only way to keep the country running will be to open the immigration floodgates.

Hit the back button and browse elsewhere

Only in the sense that the mugger who says "your money or your life" has given you a choice. In many American job markets, it's muggers all the way down.

I can't help thinking about a hidden agenda. Manufacturing moved to China so we now pay American prices for Chinese products sold by American companies. Due to the internet and business developments, consumers can sometimes buy Chinese products from Chinese companies at Chinese prices which are often a small fraction of the American price. Incidentally, Chinese products are much more likely to be repair friendly and have good availability of off the shelf 3rd party replacement parts.

I have to wonder how much of the trade war is American companies (or their wealthy owners) wanting to cut that access off wherever they can before this becomes common knowledge among American consumers.

Creating the training dataset is the *last* step. I have dozens of TB of raw data which I use to create training datasets that are only a few GB in size. Of which I'll have a large number sitting around at any point in time.

Take a translation task. I start with several hundred gigs of raw data. This inflates to a couple terabytes after I preprocess it into indexed matching pair datasets (for example, if you have an article that's published in N different languages, it becomes (N * N-1) language pairs - so, say, UN, World Bank, EU, etc multilingual document sets greatly inflate). I may have a couple different versions of this preprocessed data sitting around at any point in time. But once I have my indexed matching pair datasets, I'll weighted-sample only a relatively small subset of it - stressing higher-quality data over lower quality and trying to ensure a desired mix of languages.

But what I do is nothing compared to what these companies do. They're working with common crawl. It grows at a rate of 200-300 TB per month. But the vast majority of that isn't going to go into their dataset. It's going to be markup. Inapplicable file types. Duplicates. Junk. On and on. You have to whittle it down to the things that are actually relevant. And in your various processing stages you'll have significant duplication. Indeed, even the raw training files... I don't know about them, but I'm used to working with jsons, and that adds overhead on its own. Then during training there's various duplications created for the various processing stages - tokenization, patching with flash attention, and whatnot.

You also use a lot of disk space for your models. It's not just every version of the foundation you train (and your backups thereof) - and remember that enterprise models are hundreds of billions to trillions of FP16 parameters in their raw states - but especially the finetune. You can make a finetune in like a day or so; these can really add up.

Certainly disk space isn't as big of a cost as your GPUs and power. But it is a meaningful cost. As a hobbyist I use a RAID of 6 20TB drives and one of 2 4TB SSDs. But that's peanuts compared to what people working with common crawl and having hundreds of employees each working on their own training projects will be eating up in an enterprise environment.

Taiwan reinforces its silicon shield.

This is all to produce a peak of 240k EVs per year. Production "starts" in 2028. It takes years for a factory to hit full production. Let's be generous and say 2030.

Honda sold 1,3 million vehicles in the US alone last year - let alone all of North America, including both Canada and Mexico. If all those EVs were just for the US it'd be 18% of their sales, but for all of North America, significantly less.

In short, Honda thinks that in 2030 only maybe 1/7th to 1/8th of its North American sales will be EVs. This is a very pessimistic game plan.

Most of the people protesting technology putting people out of work do so because they know the same people saving big with the tech will fight tooth and nail to make sure we do NOT adapt the economy. They want the displaced to go die quietly somewhere that is not in their back yard. Unless/until that changes, every displaced worker brings us that much closer to an ugly social uprising.

We need to look at UNDER employment as well. Especially on the west coast, a number of the homeless are, in-fact, employed. Some full time. But because our job market is ruled by supply and demand (rather than needs), a glut in supply has resulted in employed people that can't even afford a shitty apartment.

Many of the homeless I see are doing the "thorazine shuffle", a gait that comes from years of treatment with major tranquilizers/antipsychotics. So for those, mental health issues is a good bet.

The last time I needed any sort of "advanced" support from Comcast, the person in India had no way to escalate other than mark it on the ticket and hope someone called me back in 24 hours (they did NOT).

It turned out someone assigned half of my already assigned block of static IPs to another customer.

So step one, give the AI a way to ring tier 2 at least.