In a discussion titled “Configure AI Chatbot for Trust Level 0 Access,” oppman seeks to enable new, Trust Level 0 users to access the AI chatbot, noting that private messaging (PM), required by the chatbot, is typically disabled until Trust Level 1. They consider disabling public search to incentivize login, while retaining the chatbot—which pulls from internal and external sources—for logged-in users only, and plan to use quotas to manage costs and abuse. merefield suggests hiding the search button via CSS for anonymous users as a workaround. oppman clarifies they’re using the Discourse AI plugin (installed by an intern) and asks whether switching to the Discourse Chatbot would allow Trust Level 0 access, and whether both plugins can coexist. They note their community primarily serves tech support users who often register only to ask questions—typically at Trust Level 0—and view these users as valuable due to the specialized nature of their queries. merefield confirms that enabling chatbot access for Trust Level 0 is configurable in the Discourse Chatbot plugin, with no inherent security barrier, and highlights its advanced quota system for monitoring PM activity—ideal for managing new users.
How can I enable new users to access the AI chatbot at trust level 0? I believe that the ai chatbot requires PM access and the PM is disabled until trust level 1.
I am considering disabling search and just relying on the chatbot.
Here’s my thinking:
chatbot gets both internal and external sources. I can continue to tune the chatbot
chatbot is only accessible to logged in users, which I want
I am considering disabling search in order to reward login
I plan to post a top banner on the community to let registered and non-registered users know that there are more features if they log in, including the chatbot
I have quotas on the LLM enabled and can control costs and abuse that way.
Is there a way to disable the search feature for users not logged in? The main reason I want to disable it is to reward logging in.
I am using the Discourse AI plugin, not the Discourse Chatbot. This is only because I am using the tool that an intern installed. I think the intern just installed the first plugin they encountered. They did not test multiple plugins.
If I use the Discourse Chatbot, can I enable Chatbot access for trust level 0? Is there a fundamental security reason that PM with Chatbot is disabled for trust level 0?
Can I run the Discourse AI plugin and the Discourse Chatbot at the same time? Or, do I need to dump the current Discourse AI plugin and move all-in with the Discourse Chatbot?
Perhaps like many other people, there’s a large portion of our site that is “tech support” type of questions. In this scenario, people will come to the site and search within logging in. This is normal behavior. Only when they have a question do they register so that they can post.
I would like to add the chatbot as a value-add to registered users. However, most will be at Trust Level 0 when they ask their question. I imagine that this is the same for most technical support forums. In our case, any user that asks any question is generally fairly valuable since the questions are generally about building custom software for an expensive piece of hardware.