Comment by PorciiVorbesc
5 hours ago
OK, but what's the sample size of that and who's measuring it? I never heard of that blog or took part in that poll. So how is that blog link the yardstick but mine is not?
5 hours ago
OK, but what's the sample size of that and who's measuring it? I never heard of that blog or took part in that poll. So how is that blog link the yardstick but mine is not?
Because that blog actually talked to real people and did real work, and yours is just a hallucinated list from one of the me-too LLM vendors
>Because that blog actually talked to real people and did real wor
How did you verify that those people from the blog are "real"?
I also talked to real people for my own data, case in point, I asked my mom and dad which linux DE is most used, and the results came out different. Which "real people" are the ones representative for the ground truth of Linux DE sahre?
> and yours is just a hallucinated list from one of the me-too LLM vendors
How do you know it's hallucinated? Ask the LLM the population of your country? Is the answer mostly accurate or is it hallucinated in an inaccurate way?
Aren't LLMs just outputting the highest statistical probability from the aggregate of their scraped data, which in this case would be including opinions on Reddit, and every website and blog on the entire internet (including that random one posted by badc0ffee) on the Linux DE uusage topic, making it a more accurate real-world representation than just a single random blog?
You can call it "hallucinated" if you want, but that doesn't mean it's not accurate. I asked for proof that my answer was inaccurate, not that it was "hallucinated", those are two different things, and your argument didn't prove it was inaccurate nor did it prove it was hallucinated. Would you like to try again?