That may give you a handle on the quality of his thinking, but not on whether he does. Consider all the arguments about zimboes, etc. And I believe that my dog thinks, but he can't express his thoughts in words.

I really think most of this discussion is because "think" is not a well-defined term.

Different people have different definitions of think. This argument is about language, not about reality.

It probably is, but that depends on you you understand the words. We say that airplanes can fly, but we rarely say that submarines can swim, except in recognized metaphor.

OTOH, computers have been called "thinking machines" since the 1950's, perhaps earlier. This implies that what they are doing is thinking in at least some meanings of the word.

That said, it's also clear that LLMs don't think the same way we do. So people who use more constrained definitions properly feel that it doesn't mean what *they* mean by thinking. (And don't realize that they're arguing about grammar rather than about reality.)

Again, thinking is not well-defined. Whether an LLM can think or not depends on your precise meaning. How do you know whether your brother can think? You observe him and extend your belief that he's similar to you in certain ways that you can observe onto various ways that you cannot observe...like thinking. All you can observe is actions, not thoughts.

I question whether "believes" is reasonable to use on current LLMs. At least not without strong "guard rails". OTOH, I find saying they "think" to be reasonable. A lot depends on the precise definitions you use for those terms. To me "belief" implies an emotional commitment, that I believe current LLMs lack.

"Think" is not a well-defined term. Whether or not an AI can think depends on the precise definition of "think" that you are using.

One of the common definitions of "think" is, roughly, act in a way that causes me to attribute thinking to it...which is a rather circular definition. In attributing "though" to something, it's almost always a matter of projecting myself into the acting entity. If a dog does something that satisfied it's goal, Itend to say that it has thought about it and figured out how to do it, but that doesn't mean I think it used math. So in one sense, when you say "LLMs can't think" what you're saying is that you can't project yourself into an LLM.

Some of Descartes is rigorous, but lots of it isn't. Similarly for most of the others. The exceptions have nothing rigorous or well-defined.
Actually, the same is true of Turing. And Hawking.

Rigorous thinking NECESSARILY rests on a basis that is not justified. In geometry those are called axioms. In logic, rules of inference. And current science uses the ideas of those ancient Greeks as a starting place. But it's highly questionable that they have anything to tell us that hasn't already been included, and they had a lot of silly ideas.

To justify that assertion you must have a very specific definition of what "reasoning" entails. Would you care to share it?

In the definition of reasoning that I use, e.g., alpha-beta pruning would count as reasoning. Clearly you disagree.

OTOH, being supported by a widely used OS might well help them to *become* more widely used.

Debian supports lots of systems that aren't widely used. It's even slow to drop systems that nobody appears to use. (But it does, because testing updates for those systems is difficult.)

I know it's a radical concept, but maybe CBS fired Corbett because they were losing large sums of money. The show costs 40 million USD a year with Corbert's salary at 20 million. The fact is that there are some large YouTubers that get more viewers than Corbett does now.

Maybe if they made Corbert's salary dependent on the number of viewers he had (like YouTube does) he might get back to doing comedy again.

But they just did not fire Corbett. They cancelled the whole show. There is no conspiracy other than basic economics.

To be fair, it sounds as if they haven't figured out a decent patch. I mean when they're recommend that users make the systems safe by unplugging them, it doesn't sound like they've got a better idea.

Motivated people with lots of external support tend to win. Remember Vietnam. (I suppose one could reasonably argue that the North Vietnamese were freer than the South Vietnamese, but that's not the opinion the US press ever had.)

According to the summary you get 10X as many points for getting a soldier to surrender as for killing him. But it's not genocide to kill someone who's physically attacking you. Not even if you had the alternative of capturing him (which isn't usual).

Sorry I can't point you at a reference, but thing is that the mitochondrial environment is a really bad place for DNA to live, so over evolutionary time some of the bacterial DNA moved into the cell nucleus. Mitochondria is now an "obligate parasite", though parasite is *really* the wrong term. (I can't think of the term for obligate symbiote.)

OTOH, I'm talking about the function rather then the physical pieces. This is probably similar how some of our DNA "moved into" the plants that we eat, so now we are dependent on them for vitamin C. But the result is that much of the DNA controlling the mitochondria now resides in the cell nucleus.

Very few businesses that are involved in IT in any way have anything remotely close to decent security.

Basically, they need to reintroduce the US' Internet Czar, who should have meaningful authority and who should impose meaningful IT security standards. That small companies can't afford to hire security staff is irrelevant as they mostly either work in the cloud using SAAS, at which point their provider should be handling all the security. If you want to roll your own, then you should accept the burden of paying for adequate security. Minimum standards apply to just about everything else in life, and I'd rate getting IT security right just a little bit more important than getting cars to not roll over (you can usually survive a roll) or preventing toasters from spontaneously combusting (you can park electrical appliances away from flammable stuff).

You can avoid catastrophes with defective appliances but you can't avoid catastrophes with defective IT systems.