How harmful does a chatbot need to be for people to stop using it? How much do you yourself need to benefit to choose to ignore the harm done to others?

When I left Facebook some years ago I emphasised how it was the (very late) acknowledgement of harm done to others that was the final straw.

Being complicit in the use of tools that enforce bias is an active choice. I honestly don't think many people think of this as being complicit – but I believe future historians won't let anyone off easy.

What can matter in the moment - and for your future wellbeing - is living true to your values. I often encourage my students to start there, because we so rarely are given time and space to validate and reflect on our own values.

We won't always get it right, but that can't stop us from working to get it right more often.