Yeah, I’m interested to hear perspectives on this. Our approach to date has been to acknowledge the friction and try and make it as safe as possible for people to express specific concerns about how it is used in our work. We have open discussions about how we feel about AI being used in different contexts (taxing PRs, unverified answers, walls of text etc) and I wouldn’t say we have it close to nailed, but out of those discussions we have made two rules explicit:
- always label AI content
- value the time and attention of those that have to consume your content