You do not understand the situation. These US companies can be sued on other countries as well! Mind-blowing, I know...

You seem to be on drugs. Maybe fix that instead of posting insane ramblings?

No argument.

The new AI datacenters will mostly go away. There is no realistic way to keep them running. None at all. yes, they can continue to burn heaps of money for a few more years, but there is a limit to that and then things collapse.

And that is really the long and the short of it. LLM-type AI is immature and does not fulfill professional standards in most applications it is used for. Nobody knows whether that can even be fixed. Nobody knows whether it will still be around in a couple of years, because business models that actually generate revenue remain elusive.

What happens here is that an unproven technology with multiple severe problems is made mandatory to be studied and used by far to many people. That is not good at all.

Indeed. But there are plenty of useful idiots that do not get it and are easily manipulated into voting against their own best interests. And there is large population of deeply malicious voters as well that only care about "hurting the other side".

Only if you are completely clueless. That is not something I can fix for you.

That will stop in a few years and the workloads will mostly evaporate. Because either they can do it much, much cheaper and on much lower power or they will never find that sorely needed and still unknown business model that will generate enough revenue to keep LLMs going.

Are they actually going for the economically best solutions? May these be (gasp!) capitalists making capitalist decisions that make sense financially?

Naaa, that cannot be! The demented felon claims solar is the road to hell!

Interesting. As a European, mostly non-surprising. We have learned the lesion that letting hate-speech run unchecked is a very bad idea indeed. Yes, there are risks in classifying something as and then punishing hate-speech. The risks of not doing it are greater, as long as you have basic moral principle in place and a working rule-of-law. If you do not have them anymore, it stops mattering because you have bigger problems.

Note that hate-speech does not get suppressed by law. You can still make it, you just may have to face consequences. And yes, it may get removed, but only after the fact. Also note that "censorship", which so many of the less well informed like to claim is suppression of speech, i.e. things get checked _before_ they are published. We do not have that here by law. Some platforms chose to do it though, mainly because of really vile and really disruptive users. Nothing wrong with that, they are protecting their platform against users leaving.

And then you look at the real world and see that this is not happening there. Seriously, the US is 5% of the world. The Internet is 100%.

As Trump can post any crap and any lie he wants with no repercussions, Section 230 does not work anyways.

In other words, you have nothing. How pathetic. Whether you are for or against, at least you should make an actual effort, not an entirely invalid AdHominem.

R is mostly used by engineers for statistical modelling. For that purpose it does well. I have taught engineering students and they all preferred R to the Python even for simple things as soon as statistics were involved. But R is not designed as a regular programming language. They got taught R in their statistics classes. So while R is popular for some tasks, it does not really belong into this list at all. Apples and oranges.

Does not work. I just ran a study with a student on this. Most vulnerabilities are not found, including by the expensive models.