Hi Daniel,

I just wanted to ask, does using a formal tone in my responses mean I'm using AI? I’ve been trying to keep my responses clear and respectful, but I didn’t realize that could be seen as AI-generated.

Is it a problem to write in a more formal style here, or does that automatically raise concerns about using AI? I'm just trying to understand what's expected.

Thanks for your input! JESUS (talk) 20:39, 25 December 2024 (UTC)[reply]

Two (out of two I checked) of your posts came up as 90% AI-written according to the AI detector I used (which is falliable, but that's such a coincidence). Further, the subheadings that you used were distinctive of AI-generated messages. I still believe you were using AI to generate those two messages. (By contrast, this message registers 0%.) Daniel (talk) 22:22, 25 December 2024 (UTC)[reply]

Also, this. Daniel (talk) 22:23, 25 December 2024 (UTC)[reply]

What's the current guidance on blocking AI trolling, which this clearly is? BusterD (talk) 23:04, 25 December 2024 (UTC)[reply]

I don't feel good about feeding back the "tells" either ("subheadings") BusterD (talk) 23:06, 25 December 2024 (UTC)[reply]

There's no automatic prohibition in project-space, so it needs to be judged through the lens of general disruption. The use of subheadings is both a tell but also part of the disruptive nature of LLM-generated projectspace postings, so I don't think it's fair to say "you're being disruptive" but not explain how it's disruptive. Daniel (talk) 00:18, 26 December 2024 (UTC)[reply]

In response to your fairness objection, I think it's wrong to knowingly assist in calibration of such models. We don't feed trolls, we shouldn't tutor AIs. BusterD (talk) 00:49, 26 December 2024 (UTC)[reply]