While it may be correct in this instance, what people need to realize is, these current iterations of AI are just chatbots. Think Clippy, from early Microsoft products. The only difference is, these have domain knowledge of all of the internet, where Clippy's domain was very narrow.
But, with having all that knowledge, now it'll "hallucinate". That's the technical term for it. Basically, it doesn't actually know anything, it just tells you what it thinks sounds correct in that scenario, which can lead it to telling you that 9.11 is bigger than 9.9 because 11 is bigger than 9, and that the calculators that tell you otherwise are malfunctioning.
Different AI have different filters for how it can answer, but at its core, it's still just a chatbot with the knowledge of a bunch of forums and just answering how a random forum member would. Only it doesn't understand our reality and often substitutes its own.
Answers from AI should never be considered proof of anything. I mean, look at how it draws hands. We certainly wouldn't use that as proof that that's how hands look, now would we?
In this particular instance, it seems its filtering to allow pro-trump stances. Since Elon is a Trump supporter now, it makes total sense. It could be that Elon is exposing their censoring by censoring them in such a dramatic way.
Elon mentioned he and Trump had a long conversation before the public interview, and then during the interview at one point Trump said he wants to beat China at AI. Could be that Trump and Elon made a deal in their pre-interview that could have Grok be important coming up soon.
While it may be correct in this instance, what people need to realize is, these current iterations of AI are just chatbots. Think Clippy, from early Microsoft products. The only difference is, these have domain knowledge of all of the internet, where Clippy's domain was very narrow.
But, with having all that knowledge, now it'll "hallucinate". That's the technical term for it. Basically, it doesn't actually know anything, it just tells you what it thinks sounds correct in that scenario, which can lead it to telling you that 9.11 is bigger than 9.9 because 11 is bigger than 9, and that the calculators that tell you otherwise are malfunctioning.
Different AI have different filters for how it can answer, but at its core, it's still just a chatbot with the knowledge of a bunch of forums and just answering how a random forum member would. Only it doesn't understand our reality and often substitutes its own.
Answers from AI should never be considered proof of anything. I mean, look at how it draws hands. We certainly wouldn't use that as proof that that's how hands look, now would we?
In this particular instance, it seems its filtering to allow pro-trump stances. Since Elon is a Trump supporter now, it makes total sense. It could be that Elon is exposing their censoring by censoring them in such a dramatic way.
Elon mentioned he and Trump had a long conversation before the public interview, and then during the interview at one point Trump said he wants to beat China at AI. Could be that Trump and Elon made a deal in their pre-interview that could have Grok be important coming up soon.