An exact gender-swapped version for men has been submitted by a few people, instantly rejected.

Why was that?

How do you know it doesn't mean that many women just want to share something negative about their ex?

I don't quite follow why "it's real waves" is not "the explanation".

Ok, one then perhaps needs to explain when and how it stops being a wave, but so what.

Sony is not a big guy any more.

Dafuq could be the "wrong reasons" here, pretty please.

Are you using the free version?

I have a suite of "pro" LLMs at work, they are notably smarter than freebies. But they do a lot of mistakes too though.

Pretty much anything that was not previously discussed on the internet is likely to lead nowhere.

while approving 2 coal power plants a week.

Mm, rarely, but I got "I don't know" sort of answers from LLMs.

Fuck royalties and braindead encryption on HDMI, but I recall nothing bad about Displayport.

And some bright managers at M$ decided that's a luxury one cannot have with windows 11.

So, no, thank you.

EVs sell "well" in China, as one needs 1+ year in most populated places in China to get permission to drive an ICE car.

Of all EV manufacturers, even though heavily subsidized, only one, BYD was profitable.

So "free marketplace", my bottom.

RAG is not "training" anything, but rather creates a helper DB with your "additional content" and bunch of limitations attached.
It is absolutely not the same as being trained on that dataset.

This looks more like wishful thinking from Huang, rather than actual business.

His stock went down after recent earnings call, the next gen AI chip has issues, even before a possible perfect storm (AMD MI300x and what comes after it), NV is at 70% of AI market, with big flashy "fuck you, Huang" from juggernauts like Apple, who have opted for Googles specialized chips.

Wrong. Two years ago is when ChatGPT 3.5 was released and major public just started getting taste of what LLM is.

At this point hundreds of billions are poured into "AI" with majority of projects not only not having anything in production, but even not seeing any concrete path to it.

Unless a new LLM like breakthrough (it has surprised even its developers) happens and suddenly all that chipery becomes useful, this madness will crash.

AMD ensured that all major AI related frameworks (e.g. pyTorch) run natively on its hardware.