I believe the rate of error you state for chat. I use it to support my elderly memory. I know (somewhere deep down) the answers to most of what I ask it. Yet 1/3 of my conversations have me typing "no try again, I said recently" or "Black musician" or "1960s British politician" or whatever. It's better with prompting than my memory but not much. And for sure grock is much worse. And the encyclopedia of grock makes me break out in a sweat.
It's well written (certainly better than average AI slop), has lots of info of which I was unaware and doesn't have any obvious bias. I assumed it had been cribbed from Wikipedia, but if so I can't find it there or elsewhere. Any thoughts?
I believe the rate of error you state for chat. I use it to support my elderly memory. I know (somewhere deep down) the answers to most of what I ask it. Yet 1/3 of my conversations have me typing "no try again, I said recently" or "Black musician" or "1960s British politician" or whatever. It's better with prompting than my memory but not much. And for sure grock is much worse. And the encyclopedia of grock makes me break out in a sweat.
Try this experiment in both grokipedia and Wikipedia by entering the phrase “Trump Epstein Scandal” and see how Musk’s filters are working.
I'm generally in agreement with you on Musk's evil and stupidity. But I searched on my own name and came up with this article
https://grokipedia.com/page/Zombie_metaphor_in_politics
It's well written (certainly better than average AI slop), has lots of info of which I was unaware and doesn't have any obvious bias. I assumed it had been cribbed from Wikipedia, but if so I can't find it there or elsewhere. Any thoughts?
“Stupid” <—> “Evil” is not an either/or proposition.