I'm a huge fan of AI. I predict that it will help us solve previously intractable problems in science, engineering, medicine and maybe even politics.

Presumably you mean LLMs. LLMs cannot think; cannot synthesize new ideas; thus cannot solve unsolved problems. LLMs are not the path to AGI.

Europeans will absolutely get what they deserve here.

I think it's safe to say that the European style of democracy, where for some reason every single decision is closely scrutinized and can be vetoed by just about anyone, and every industry is regulated to the point were any change is essentially impossible, and new industries are killed before they can even get off the ground, has turned out to be a bad idea, and the US should immediately turn away from these kinds of policies before we follow suit and become irrelevant.

SEO killed search. If anything, AI has made it slightly useful again.

If you’ve ever tried to argue with an LLM, it quickly becomes obvious that they don’t even really know what words mean, and this article explains why. They don’t actually have any real life experiences to relate to those words. They’re not trying to understand the world, they aren’t even really exposed to the real world as a source of data, all they’re really trying to accomplish is arranging words into patterns that humans will find authentic and convincing. That’s the main thing these models are optimizing.

Yes, but we’re not counting on them to become super intelligent and solve all the worlds problems.

I've used them, I am well aware of what they are actually capable of, which is mostly expanding a small piece of text into a large one without adding anything of value. In most cases, it sounds like a highschool student trying to BS their way to meet a necessary word count without really engaging with the material. Trying to call it "superhuman" is actually hilarious.

Yeah, it seems like AI coding is totally misguided today. AI would be good for helping people with syntax, or identifying typos. Maybe you could use it to produce early demonstration versions of software to help set requirements, but using it to actually write code doesn't make sense for any kind of real product you intend to ship to customers, especially if security is any concern.

I'm not saying IA itself is not useful, but that AI generated content is not useful. There are certainly things it does well that can improve your productivity, but generating answers to questions, generating complex code, writing stories and things like that are not among them.

It’s good for a handful of things, but as a rule content that is largely AI generated is not useful. AI chatbots and ai generated answers to questions are not particularly informative and often contain glaring logical contradictions, nonsensical statements, and even factual inaccuracies. AI, in its current form, will never be able to think, make decisions, or teach students. It’s as if we created the world’s smartest insect and asked it to raise our children, when it should be summarizing wikipedia articles and filtering spam.

Eliminating the Electoral College and/or the First Past the Post election system would have been a good start. IMHO, a Parliamentary system stops the office of head of state from being a popularity contest.

How do you know Thunderbird doesn't monetize your data and charge you $9 for the privilege?

Probably not. The people that like AI slop are probably the same ones that watch shows like Ancient Aliens and a host of reality shows on the Discovery Channel in the US.

In the US at least, showing that you accrue temporary debt and pay it off improves your credit score.

Classical bits can store only one number at a time. They probably mean that that many qbits can store the amount of information in the universe simultaneously due to superposition.