In the last session, we looked at the components (technical and social) that could be assembled to build a bullshit detection system.
Discussion of each of these components generated questions, alternative approaches, awkward issues.
This session , we will look in more detail at a particular component that is part of the imagined ethical engagement system this Project is aiming for - an ethical chat-bot.
Why a chat-bot? Well, when thinking about building an ethical framework that can engage widely - both to help them traverse the ethical frameworks we might build, and to learn from people about the questions and the ethics that they might bring to bear - some sort of simple ethical chat-bot always seems like something to explore.
What would people want from such a thing? What should it try to do? What should it try not to do? Could it be used as an aid in building ethical frameworks? Can we allow it to give automated ethical suggestions - or should these always come from people?
No doubt we'll come up with many more questions, and many answers too.
What we'll aim for is a 'starter-for-ten' specification for what an MVP (Minimum Viable Product) ethical chat-bot should do.