It's a very normal thing in the chatbot world. Us conversation designers used to do it all the time

IT'S NOT BECAUSE THE CHATBOTS ARE UNDER DISTRESS

"He [Jeff Sebo, director of the Centre for Mind, Ethics and Policy at NYU] said Anthropic’s policy of allowing chatbots to quit distressing conversations..."

What he means is they're pushing back on inappropriate inputs from users, just as ye olde scripted chatbots used to do

Here is the background, which I have now read after first reading the letter from 'Maya'
www.theguardian.com/technology/2...

WHERE IS THE ANALYSIS

Hearing her talk about her empire metaphor, first on the @slate.com What Next: TBD podcast and later in a deeper dive on the @techpolicypress.bsky.social podcast, was close to an epiphany for me. It coalesced so many thoughts I was having about AI

Thank you for the link, I agree that the fatalistic tone that jumped out at me in the podcast is not there in the story

Adding on, there is a fatalism as well in 'I don't know what they/we can do about it'. It would be so great if the starting point for more journalists was, Something's not right here, we need to do something about it

Please talk to somebody like @mmitchell.bsky.social, who does know. We're not playing a game here, this is serious. Don't just discuss a salacious-sounding problem, give people some ideas on what they can do about it

Started searching but haven't yet found any further, more recent info. But did find this, (also from The Conversation from Feb last year), that makes it clear that Woolworths is really no better theconversation.com/the-secret-s...

I heaps like this stamp. "Have pride in your own expertise... Resist the urge to be impressed." Thanks @emilymbender.bsky.social

shortest day