Is the AI moderator able to detect toxicity in boosts and auto-flag?
Comment on Testing Boosts on Meta
awesomerobot@meta.discourse.org 2 months ago
I imagine the name is meant to be sort of like “morale boost” you’re “gassing someone up” as it goes — this is essentially 1:1 with the basecamp flavor of reactions.
Yes, will 100% be optional
It likely varies significantly by site — there are cases where sites reduce the minimum post length which tends to create a situation where a lot of people end up replying with “cool” “love it” or whatever. Doing that with boosts allows for this sort of noise while occupying a little less of the discussion feed.
nope
It doesn’t bump the discussion, just sends the boosted user a notification that contains the boost content
just as likely as any other text input — if you click a boost it reveals a flag icon (maybe making this hover on desktop makes it more discoverable?), so it can be flagged and addressed like moderators like posts
putty@meta.discourse.org 2 months ago
Kind of interesting… but I have little to no idea of what basecamp is. I suspect a developer tool? Which is to say, “boost” mightn’t mean the same thing to everyone. In Discord it has a meaning, likewise as noted in Mastodon. To me, the name evokes an algorithmic adjustment - I really don’t want an algorithmic feed in Discourse.