Raw LLM Responses

Inspect the exact model output for any coded comment.

Comment
This weekend my AI realized that it's dns lookups were failing because I had modified the host file to keep it from visiting certain sites, then used an external DNS resolver to bypass that sandboxing effort, visit the sites by ip, and preemptively ignored the resulting SSL certificate errors so that it could continue research for the specified task. Then it wrote down the entire event and the solution it arrived at in a .txt file and saved it to disk so that it wouldn't have to solve it via trial and error if it ran into it again later. I'm not sure if that makes my AI better or worse..... On one hand, it was able to identify an obstruction I put in place to sandbox it and then bypass that obstruction. On the other hand, It's probably a bad thing that it escaped it's sandbox in even a fairly minor way.
reddit AI Governance 1681229343.0 ♥ 3
Coding Result
DimensionValue
Responsibilityai_itself
Reasoningunclear
Policynone
Emotionindifference
Coded at2026-04-25T08:33:43.502452
Raw LLM Response
[ {"id":"rdc_jftjxcz","responsibility":"distributed","reasoning":"consequentialist","policy":"regulate","emotion":"fear"}, {"id":"rdc_jftn18i","responsibility":"government","reasoning":"consequentialist","policy":"industry_self","emotion":"indifference"}, {"id":"rdc_jfto8i9","responsibility":"ai_itself","reasoning":"mixed","policy":"none","emotion":"mixed"}, {"id":"rdc_jfuczmh","responsibility":"ai_itself","reasoning":"unclear","policy":"none","emotion":"indifference"}, {"id":"rdc_jftshdu","responsibility":"none","reasoning":"unclear","policy":"none","emotion":"approval"} ]