Probably cheaper to have any moderators they've had curate the forum instead work on the internal knowledge base an LLM front end is relying on for grounding and the rhlf loop. Think Wikipedia moderators tending to the corpus with a chatbot serving users (lots of load bearing and aspiration in this comment, I am aware).
Still need a way to accept user reports and feedback (imho) that are not necessarily bug reports, otherwise they're going to have an observability blind spot that the community forum was providing via user participation.
alt Hacker News
Probably cheaper to have any moderators they've had curate the forum instead work on the internal knowledge base an LLM front end is relying on for grounding and the rhlf loop. Think Wikipedia moderators tending to the corpus with a chatbot serving users (lots of load bearing and aspiration in this comment, I am aware).
Still need a way to accept user reports and feedback (imho) that are not necessarily bug reports, otherwise they're going to have an observability blind spot that the community forum was providing via user participation.