Clearly. Cost of doing business when you put a Mafia-style government into place. Of course, everything goes to crap as well, but apparently the voter wants that.

Indeed. No idea why he still tries getting richer at his age though. He basically threw away his whole live on an obsession of being something he is not. I would not care much, except for the massive, massive damage, the rape, war-crimes, scamming, climate-change-denying, etc. is doing to everybody else.

Mythos has done some impressive stunts, but stunts is all they are. And not-so-smart people let themselves be blinded.

For example, of the hyped "271" FF bugs (which were actually 3), sound coding practices and "gcc -f analyzer" (or other tools) would have found at least 2, the "use-after-free" ones, which should not be happening in competent software creation anyways.

The thing that is happening here is that money that should have been spent on software quality was not spent. Then AI comes along and the money gets spend indirectly via its use. Actually competent tool use and dev practices would have done more than the LLM and at a much lower price.

I completely agree. But I would like to point out that "clever marketing" is used on people that are basically defenseless because they want to believe, not see reality. The same happened before in all AI hypes, just not on this scale. And it never works out to more than a gradual improvement, maybe 1...5% of the grand promises.

I agree. This is to make LLM tech look like much, much more than it is. In actual reality, it can, for example, have a place in software dev and things like Mythos can even, occasionally, with low reliability, find instances were the devs really screwed up (like the hyped FF bugs, which were 3, not "271" and two were "user after free", which even gcc finds with -f analyzer and, incidentally, only one of the 3 had a "high" rating). That said, I recently talked to some people that try to do it and the process of generating production quality code with LLM support is really, really complex and high effort. Complex enough and high effort enough that it probably is overall not cost effective and costs you money.

The same is probably true for other instances of LLNs "proving useful": Had the same money been given to actual experts, the results would have been comparable or better. And that does not take into account that when LLMs do things, educating experts becomes much harder, less human expertise gets documented (which makes LLM training much harder) and things will probably become stagnant, instead of moving slowly as they currently do.

My take is the LLM craze is yet another instances of businesses and C-levels being incapable of making long-term plans and being all focused on the next quarter. That is not how total cost calculations work. It does nicely show that business graduate education is crap, though.

My grandmother lived near an airbase, and every day, there were at least two supersonic aircraft in the skies. When they were flying above us, you could not understand the other guy talking. And we have to consider the distance in the line of sight. A starting aircraft on an airbase on the other side of the valley might be 10 miles away from you. A starting airplane right above you is quite different.

I wonder why most other nations also had a ban of non-military supersonic flight in place, nations with supersonic passenger jets, a.k.a. France and the United Kingdom, and those without, like Brasil or Germany.

Supersonic flight is incredibly noisy, and you don't want it above you.

That matches my stance. Unfortunately, it often is the second kind.

Indeed. The question is why now?

Context matters. Read more than just the three words _and_ realize that the first three words were a relative statement to the previous comments.

Seriously, this is not kindergarten.

Yes. I really do not get the anti-AC sentiments here in Europe. Yes, 20 years ago, there was zero need for AC, but that has very obviously changed. The only calculation I did was how long I have to work to finance the AC electricity and AC itself for running it for a day. Came out to less than 10 minutes and that, I think, is entirely worthwhile. Also, obviously, solar brings in more power when the sun blazes, so....

"We"? fafalone discussed Supergirl (in which Milly Alcock is the lead actress) because you were first talking about Supergirl.

TFS: "... Supergirl ... Supergirl ... Supergirl ... Supergirl ..."
angel'o'sphere: "Is the main actress 'barely legal' As in Sexy, young like 16 or so? [shift to summer blockbusters like Transformers to make a point]"
fafalone: "The lead actress is Milly Alcock who's 26. You ok there?"
angel'o'sphere: "[I only remember Transformers, so I don't understand the flow of the discussion]"
Milly Alcock is not in any Transformers movies or television shows as actress or voice. That should have been clue #1 that fafalone was talking about the first thing you mentioned: the Supergirl movie which is the primary subject of TFA and mentioned multiple times in TFS.

Nope. And where did I say "not a problem"? Please provide a citation.

While it does cut down on time, there is a massive problem here: Experts need to maintain their expertise. With the use of LLMs to replace them, that does not happen anymore or happens less. As we clearly already have to few experts, any LLM use is too much.

What happens long-term is that we will not have the experts that create the data the LLM was trained on anymore and no updates happen. This then leads to complete stagnation of the discipline, which is much, much worse than slow discoveries. Typical "next quarter will be great!" thinking at work, that overlooks that short-term optimization quite often leads to long-term catastrophe.

So, not "good news" at all. Exceptionally bad news instead. Although it will take a while to become obvious.

For some of them, yes. But why the denial in the others?