How about "smart assistants should not encourage people to do stuff based on what is blindly scraped on the internet"? Seriously, how could they not have seen

How about "smart assistants should not encourage people to do stuff based on what is blindly scraped on the internet"? Seriously, how could they not have seen that it's a bad idea to have a "challenge" prompt say random stuff from the internet? I get that if I ask a question like "how high is mount everest" the answer can be "According to wikipedia it is 8,848 meters high" but doing the same for "give me a challenge" is pretty stupid.