I am regularly mean to it. Because it will do the dumbest fucking shit and I’m not letting it slide.
Do people at Anthropic think claude is their girlfriend?
When I read about this on Lemmy for the first time I thought this was a joke.
Looks like a pretty straightforward publicity stunt to me, low cost - high controversy.
Does this indicate that they’re training it on the users’ words, and they don’t want it to learn bad habits?
Maybe we’re the ones being trained all along‽
Probably not, enforcement of this is likely automated and they could just as easily filter out the chatlogs they don’t want to use for training instead of telling users what to do.
Having this in place, even automated, can discourage user from exhibiting bad habit that cannot be automatically detected.
