uz
Feedback
String's18+RapidFireMemes💦

String's18+RapidFireMemes💦

Yopiq kanal

🚩 Channel was restricted by Telegram

Ko'proq ko'rsatish
1 252
Obunachilar
-124 soatlar
-67 kunlar
-1330 kunlar
Postlar arxiv
Case in point ^^^

I dispise that they have programmed chat bots to mimic first person speech. This makes it feel natural to assume they have intentionally. Like, we anthropomorphize spotty mechanical devices for fun like they have personality, but we understand they have their own make-believe interests (and we're allowed to treat their opinions as secondary to other humans, for one). Also the sourcing thing it does now is anti-intellectual trash. Okay, you would have gotten an F in school if you cited sources like that. Like, there is so much on wiki that it's probably true there is something related there you can read. We don't accept that as a source. LLMs can't cite because research isn't what it does. The data it uses was already baked in. I'm guessing grok is merely programmed to add "mY sOuRcE iS tHe InTeRnEt" at the end and point to vague websites at random. Citations include the authors, websites, book names, and dates so that people can find them. A popular thing like TW will influence many into thinking this is acceptable. LLMs generate text from math, not research, which is why you'll never see is them quoteing ppl. Their output is probabilistic, not conceptual. It is foul to make ppl think an LLM researches by adding in website names. The world is being introduced to AI in such a misleading way. The majority of ppl are focused on how factual the output is and not understanding the method.

photo content

Yeah, I am.
Yeah, I am.

photo content

D(eer)NA 🦌  🍓  🦌🍓   🍓  🍓🦌 🍓  🦌 🍓  🦌 🍓  🦌 🍓 🦌  🍓🦌   🦌  🦌🍓 🦌 🍓 🦌  🍓 🦌   🍓 🦌  🍓 🦌 🍓  🦌🍓   🍓  🍓🦌 🍓 🦌 🍓  🦌

photo content

photo content

String's18+RapidFireMemes💦 - Telegram kanali statistikasi va tahlili