No, it's not remotely OK. But it is certainly inevitable. These tools are already looking like they're going to bring all sorts of terrible things to us, though, so bringing more ad crap is comparatively a small problem.

"ChatGPT, how can bumblebees fly?"

"What a wonderful question, it certainly speaks to your superior intellect. Now let me tell you how dumbcorp revolutionized male shaving by rethinking the concept of the blade..."

No

Sounds like a great use for a local LLM to strip ads from the output of the ad-infested LLM.

It’s actually extremely bad. You won’t know about the ads this way. It’s vastly worse than most already terrible advertising.