And don't mistreat peoples public data, and expect them to like or support you.

Unfortunately “mistreat” is highly subjective. No matter what you do someone will be angry at you. I once got yelled at for taking a picture on a public beach where there was some family picnicking maybe 50 meters away. I think I was reasonable, the gentleman most emphatically did not.

Control what you can control. If you object to being a small data point in someone else’s documentation of a public experience, don’t put yourself in that situation.

Don't disagree, just saying it goes both ways. In your case you didn't care (not judging, I probably wouldn't either) about the opinions of the randos at the beach. However, in business, reputation does matter.

I'm fine with public records of stuff I have posted, I'd not be fine if you were impersonating me, nor I am if it's a piece of software doing that.

Can we take the ethics of AI seriously? I feel it's about time.

>Can we take the ethics of AI seriously?

If you're not suggesting a law to do so, then no. 35+ years of using the internet tells me that ethics is not included, nor at this point in the game should be expected.

Agreed, role playing as real people is unacceptable for both AI and other real people.

“Tell me what Bernie Sanders might say about…” is fine, so long as the response is in the form “based on his past statements”. “Pretend to be Bernie Sanders and talk about” is not ok to prompt, nor ok for the model to respond to with an impersonation.