The discussion focuses on user feedback for Ask Discourse, highlighting both its utility and technical issues. Lee_Ars and Ed_S initially praise the tool for summarizing Meta search results and providing helpful answers. However, significant concerns were raised regarding the AI’s reliability. one1 reported that the bot provided generic, non-Discourse-specific advice and generated links to non-existent posts, a problem echoed by Canapin. one1 suggested testing against a finite list of actual posts to prevent hallucinations. mcwumbly advised users to use the negative feedback button to help tune the model. Ed_S later reported the AI outputting incoherent text, which sam attributed to a site upgrade; the issue was resolved after the upgrade, though minor markup errors persisted (Ed_S).
Other technical issues included rate-limiting errors for matthewruzzi, which HAWK and awesomerobot identified as standard Discourse spam filters rather than Ask Discourse-specific bugs. UI feedback from twofoursixeight noted that the Ask Discourse website sidebar contained unused links (like
בדרך כלל אני מוצא ש-Ask עוזר למדי. אבל היום הוא סיפק שני קישורים לפוסטים שאני לא יכול לגשת אליהם. אני מניח שהם קוטלגו לקטגוריה עם גישה מוגבלת - חבל, כאן, כי הם נראו כאילו הם עשויים היו להיות שימושיים.
I asked it, “when a moderator rejects a flag, does the post get unhidden” and it gave me general information about non-Discourse forums. Shouldn’t all the advice be assumed to be about discourse? What is the point of the tool if it isn’t oriented that way?
מצאתי שזה הרבה יותר יעיל מאשר לחפש תשובות במטא — אני יכול להטיל עליו שאילתה אחת בשפה טבעית (כמו “איך אני גורם לדיסקורס לעשות \[דבר\]?”) ולקבל תשובה המורכבת מחלקים מסוכמים מאותם 3-4 חיפושי מטא שהייתי מוצא בסופו של דבר, אבל בלי שאצטרך להשקיע את הזמן כדי למצוא אותם. אני בהחלט מתכנן להמשיך להשתמש בו כשיש לי שאלות מסוג “איך אני עושה” או “האם אפשר” בנוגע לדיסקורס.
I was wrong before when I said those messages don’t apply on ask.discourse— they still do because it is powered by Discourse. You are being hit by the normal spam rate limiting.
Usually the “too similar” message only happens when the post you’re trying to make is identical to an earlier post… I also notice that you submitted two identical flags, which might have been the cause of the “creating topics too quickly” error… perhaps some messages were attempting to send twice in rapid succession due to a flaky internet connection? this could potentially explain both issues
Is there any way to tell the model that queries for the DataExplorer should never end with a ;. I know it seems like a small thing, but when I have to go through several iterations I’m constantly forgetting to remove it.
The “topics”, “my posts”, “categories”, “my messages”, and “filter” do not serve any use on the site either, as the Ask Discourse site is mainly for Discourse AI chat and not for making topics. Those pages are empty for 99% of people and should redirect back to the AI bot if you are a normal member. Otherwise, the website functions well for asking an AI questions.
It’s oddly easy, let me show you through screen recording.
There are a couple ways to access it, as the sidebar isn’t always defaulting to the AI bot one especially in user settings. Clicking all questions, fonts and appearance, or account settings uses a different sidebar than the intended one, that lets you access the unused topics section or unused screens
Note to get back to the ai chat, you click AI bot or the logo.
button.
