They tell you stuff similar to the training corpus that the people tagging it want to hear.
It’s close to what you said, but the difference is actually important some times. In particular this one seems to not have been exposed to “corporate speech” while training.
The chatbots tell you what you want to hear.
Don’t forget that.
I don’t know on this one, with how shit Musk’s recent projects have been, this one might be broken enough to be more right than not
They tell you stuff similar to the training corpus that the people tagging it want to hear.
It’s close to what you said, but the difference is actually important some times. In particular this one seems to not have been exposed to “corporate speech” while training.
This should be the only comment on anything grok related.
But they all for this obvious fake.