I haven’t used the AI much yet, although it’s one reason why I went with Discourse. Thanks for the link to ask.discourse.com. I will play with that.
As to a learner’s mode using an AI-powered bot, that sounds useful. Site-awareness seems essential, but there are different perspectives of learners.
Are they staff, who need to understand the Discourse platform and manage users and content? Or are they users, who will be focused mostly on the content and other users and will be resistant (probably) to having to learn anything more than they have to about the platform/tool itself. The more friction they encounter, the less likely they’ll be to participate or stick around. (Or maybe they’ll just be intimidated…)
There is some evidence out there that naive users might be able to use an AI more easily than learning the ins and outs of a complex tool. The latest announcement from The Browser Company re: Arc might be a case in point.
tldr: Arc will be maintained for existing geeky users, but it’s not going to scale, so they’re starting a whole new project for non-techie “friends and family.”
As a new admin, I’m in (geeky) learner’s mode for Discourse as a platform/tool, and as a new user here on Meta (who does actually want to learn the platform). My real focus is elsewhere, though, i.e. setting up my own forum, including making a lot of decisions about my unique configuration, my unique starting set of content, and my new and future members. It’s a lot.
My biggest fear at this point is making decisions that will be hard to back out of later. I fully expect to experiment and to evolve both my set-up and content over time, but some decisions seem stickier than others. I’m particularly conscious of what my earliest users will need to make the project a success in the long run, ultimately. I want to get off on the right foot.
I went through discobot’s standard tutorial (not the advanced one yet), and I’ve been poring over docs and wiki here on Meta. There’s a lot of really advanced content intermixed with “what you really need to know to get started,” and it can be hard to tell the difference.
Maybe a way to think about an AI bot for Meta – and as a tool for admins to use on their own forums – would be to consider what kind of persona with what kind of information the bot/AI should have to be helpful for different needs, for example:
- admins at various stages of the game, with various concerns front of mind
- new or experienced users, also at various stages in the evolution of the community they’re joining