Raw LLM Responses

Inspect the exact model output for any coded comment.

Look up by comment ID
Random samples — click to inspect
G Uh you’re misunderstanding, you wouldn’t be someone “watching someone’s head get… ytr_UgzvSH8z-… G You can replace WORKERS with AI, but you can never replace CUSTOMERS with AI. An… ytc_UgxWcnuRm… G Exactly.. human radiologist make more error than AI.. and AI is gonna exponentia… ytr_UgxDtenlY… G 14:28 --- As a disabled artist who can barely get out of bed most days without m… ytc_UgxdCKkgx… G I think it’s less Terminator and Skynet, and more lack of basic human decency, a… ytc_Ugyrn1U0G… G I hope the writers get the compensation they deserve. The cgi and special effec… ytc_Ugzp_dAxd… G That's why rayban developed those perv glasses. Basically ai can feel what it fe… ytr_Ugztite_u… G Love the feedback loop. AI Creates Content > Content Creators, like this channel… ytc_UgyulCFpT…
Comment
It doesn't improve itself, it optimizes another tiny model by running automated engineering loops. Right now it's constrained by the main model's capabilities as Karpathy said, but it's easy to see how that automated engineering will scale up to actual research too in the future.
reddit AI Moral Status 1773250766.0 ♥ 6
Coding Result
DimensionValue
Responsibilitynone
Reasoningmixed
Policynone
Emotionindifference
Coded at2026-04-25T08:33:43.502452
Raw LLM Response
[ {"id":"rdc_o9wx8yo","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"indifference"}, {"id":"rdc_o9xyhxg","responsibility":"company","reasoning":"deontological","policy":"liability","emotion":"outrage"}, {"id":"rdc_o9w2cei","responsibility":"none","reasoning":"mixed","policy":"none","emotion":"indifference"}, {"id":"rdc_o9wack6","responsibility":"developer","reasoning":"virtue","policy":"none","emotion":"approval"}, {"id":"rdc_o9wsyba","responsibility":"company","reasoning":"mixed","policy":"regulate","emotion":"mixed"} ]