Raw LLM Responses

Inspect the exact model output for any coded comment.

Comment
the part that gets underplayed is the feedback loop. earlier influence ops flooded zones with static content and hoped something stuck. the new version can run thousands of message variants in parallel, score engagement, and iterate on what actually moves a specific community. cambridge analytica was guessing by comparison, this is more like continuous ab testing on political persuasion. the coordination efficiency is the real threat, not just the volume.
reddit Viral AI Reaction 1777058655.0 ♥ 2
Coding Result
DimensionValue
Responsibilityunclear
Reasoningunclear
Policyunclear
Emotionfear
Coded at2026-04-25T08:06:44.921194
Raw LLM Response
[ {"id":"rdc_oi299g0","responsibility":"none","reasoning":"unclear","policy":"unclear","emotion":"outrage"}, {"id":"rdc_oi2gix5","responsibility":"unclear","reasoning":"unclear","policy":"unclear","emotion":"outrage"}, {"id":"rdc_oi2idvt","responsibility":"developer","reasoning":"deontological","policy":"liability","emotion":"outrage"}, {"id":"rdc_oi2if5r","responsibility":"unclear","reasoning":"consequentialist","policy":"unclear","emotion":"fear"}, {"id":"rdc_oi2j0c6","responsibility":"unclear","reasoning":"unclear","policy":"unclear","emotion":"mixed"} ]