(C) Daily Kos
This story was originally published by Daily Kos and is unaltered.
. . . . . . . . . .



Time to Fight Back Against Trolls — With AI That Understands Their Behavior [1]

['This Content Is Not Subject To Review Daily Kos Staff Prior To Publication.']

Date: 2025-05-13

If you've spent any time in the comment sections of Daily Kos — especially on front-page diaries or hot-button issues — you've probably seen it: the subtle derailment, the endless “Just Asking Questions,” the oddly persistent contrarians who show up not to engage, but to exhaust.



We all know trolls exist. But what if I told you there's a way to identify them based on how they behave, not just what they say?



I’ve been exploring this issue with the help of ChatGPT, and we came across a recent article from The Conversation titled:

👉 Some Reddit users just love to disagree, new AI-powered troll-spotting algorithm finds

(https://theconversation.com/some-reddit-users-just-love-to-disagree-new-ai-powered-troll-spotting-algorithm-finds-255879)



It describes a powerful idea: trolls often give themselves away not by slurs or spam, but by patterns — patterns that AI can now detect. A group of researchers used inverse reinforcement learning to analyze Reddit users who routinely sought out disagreement. They weren’t banned, they weren’t obviously abusive — but they made communities worse.



Sound familiar?

🧠 What We Could Do at Daily Kos

In conversation with ChatGPT, I posed this question:

“Could you (ChatGPT) be contracted by Daily Kos to examine all its comments to find a population of trolls?”



The answer was yes — and the method is surprisingly effective.



Here’s how it might work:

- Track who replies to whom, and how.

- Spot users who consistently argue without contributing.

- Flag patterns of subtle bad-faith engagement, like concern trolling, deflection, or coordinated brigading.

- Use this analysis to help moderators, not replace them — think of it as radar, not missiles.



This wouldn’t target people with different opinions — it would spot those who come here only to provoke, derail, and waste our energy.

📊 Why It Matters

Daily Kos is one of the most important progressive platforms in the U.S. But our community strength depends on trust. Trolls erode that — not with one comment, but with a thousand micro-cuts.



If the right-wing can flood social media with bots and bad actors, we should respond with smart, ethical tools that defend the conversation, not suppress it.

🤝 What I’m Asking

- Would you support using AI tools to detect bad-faith actors in the comments?

- Would you like to see a trial version that maps behavioral patterns (not content) over time?

- If you’re a moderator, would this kind of radar help you?



I’m asking the Daily Kos team and the broader community to consider this seriously. If we want to keep this space strong, we have to outthink the trolls — not just argue with them.



Let me know in the comments — and if you want to see the full conversation I had with ChatGPT, just ask. I’ll share the details.

[END]
---
[1] Url: https://www.dailykos.com/stories/2025/5/13/2322113/-Time-to-Fight-Back-Against-Trolls-With-AI-That-Understands-Their-Behavior?pm_campaign=front_page&pm_source=more_community&pm_medium=web

Published and (C) by Daily Kos
Content appears here under this condition or license: Site content may be used for any purpose without permission unless otherwise specified.

via Magical.Fish Gopher News Feeds:
gopher://magical.fish/1/feeds/news/dailykos/

You are viewing proxied material from magical.fish. The copyright of proxied material belongs to its original authors. Any comments or complaints in relation to proxied material should be directed to the original authors of the content concerned. Please see the disclaimer for more details.