Reddit users struggle to spot AI bots as comment quality gets eerily human
Borkato · reddit · 2026-11-01
A Redditor documents comments from one account that look increasingly bot-written: highly specific Stable Diffusion training advice, natural phrasing without telltale repetition, even 'human' touches like owning a golden retriever. The poster, who used to spot bots instantly, now hesitates—highlighting how AI-generated content is penetrating technical communities and how detection cues (hidden history, tone consistency) are eroding. The thread also asks how to make local chatbots sound less robotic by comparison.
More from AGI Musings
- Yacine MTB: Robotics Hardware Will Soon Bottleneck Autonomous Progress, and He'll Help — yacineMTB · 2026-11-01
- FT's Burn-Murdoch: US productivity surge is not AI — it's a uniquely American labor pattern — jburnmurdoch · 2026-11-01
- AI will unlock the world's hidden creatives, citing MIT 40% productivity study — miilesus · 2026-11-01
- Ethan Mollick: slop reply bots now evolve to telling lame jokes for likes — emollick · 2026-11-01
- Ethan Mollick floats funding AI to unlock the Oxyrhynchus Papyri backlog — emollick · 2026-11-01
- Should science regulate its own AI use? French debate over human expertise vs state oversight — IgorCarron · 2026-11-01