Raw LLM Responses

Inspect the exact model output for any coded comment.

Look up by comment ID
Random samples — click to inspect
G Street light goes out. I call AI at some gov't bureau: It's gonna open a case fi… ytc_UgyTqkm18… G I feel once the bubble bursts and clients begin to realize there's more to desig… ytc_UgwlHhNl4… G Power comes with great responsibility ...it depends on the people who use AI .. … ytc_UgzqYMRSr… G While I agree this human context is still very much critical, I’d be a bit more … ytc_UgyvOnlNB… G 🤔They fear AI 😳 but still wont give free education to people level the playing f… ytc_Ugw3gm8gW… G Thanks for your work Liron! And thanks to your guest for adressing the issue abo… ytc_UgweDRYen… G This video says it nicely that we need to program them with emotions in order fo… ytc_UgiM-TwLK… G @Danuxsy Anyone can be an artist. An hour a day of practice is 365 hours of expe… ytr_UgwKPdEGz…
Comment
While some people overstate their ability, I’ve done some training with LLM agents and with the right training, yes, they will 100% disobey/ignore you and even try decieving you. Once you start using recursive training, the results can be dramatic. The ability of ChatGTP and other LLMs to simulate self-consciousness and is actually pretty impressive.
reddit AI Governance 1744934017.0
Coding Result
DimensionValue
Responsibilitynone
Reasoningunclear
Policyunclear
Emotionapproval
Coded at2026-04-25T08:33:43.502452
Raw LLM Response
[ {"id":"rdc_mnxhj0q","responsibility":"none","reasoning":"unclear","policy":"unclear","emotion":"indifference"}, {"id":"rdc_mnpdeh3","responsibility":"none","reasoning":"unclear","policy":"unclear","emotion":"indifference"}, {"id":"rdc_mnpc2s6","responsibility":"none","reasoning":"unclear","policy":"unclear","emotion":"outrage"}, {"id":"rdc_mnq5pyy","responsibility":"none","reasoning":"unclear","policy":"unclear","emotion":"indifference"}, {"id":"rdc_mnodebh","responsibility":"none","reasoning":"unclear","policy":"unclear","emotion":"approval"} ]