Raw LLM Responses

Inspect the exact model output for any coded comment.

Look up by comment ID
Random samples — click to inspect
G I think, this problem becomes much easier if we speak not of humans, but of pers… ytc_UgycfRkO7… G That's an interesting perspective! Sophia raises a valid point about the balance… ytr_UgzAacA1q… G We're not ready for A.I. The concept is too early for us. We should at least wa… ytc_Ugxz2Tcr9… G You need it on an international scale to work though. Otherwise, whichever count… ytr_UgxPxzApN… G AI is not magic …it’s only as good as the data it trains on. We are far from one… ytc_UgyBYWyp0… G even if you use "AI" occasionally and only for "perfect AI tasks" - you still lo… ytc_UgwUifwvw… G Next gen AI is already being trained with current gen AI. That’s the easiest wa… rdc_l9vtcgl G And Pandora's box has been opened. The 'singularity' will morph into an electron… ytc_Ugyyw77kb…
Comment
If you have a false negative rate of 0% and a false positive rate of 0.01% (99.9% accurate) then you seem like you have a very good algorithm. The problem is that applying this to a VERY large pool that is known to be filled with people without whatever trait you are looking for is that 0.01% of that pool is a LOT of people. If you're looking across the entire US population for a single person that committed a crime this will return: True Positives: 1 \* 100% = 1 person False Positives: 331,449,280 \* 0.1% = 331,449 people ​ So now your criminal is actually only 0.0003% of your "guilty" pool.
reddit AI Harm Incident 1626260612.0 ♥ 30
Coding Result
DimensionValue
Responsibilitydeveloper
Reasoningutilitarian
Policyunclear
Emotionmixed
Coded at2026-04-25T08:33:43.502452
Raw LLM Response
[ {"id":"rdc_h5415fw","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"resignation"}, {"id":"rdc_h54t138","responsibility":"user","reasoning":"consequentialist","policy":"regulate","emotion":"outrage"}, {"id":"rdc_h5429ez","responsibility":"government","reasoning":"deontological","policy":"none","emotion":"indifference"}, {"id":"rdc_h54hw5v","responsibility":"developer","reasoning":"consequentialist","policy":"liability","emotion":"fear"}, {"id":"rdc_h553r3q","responsibility":"developer","reasoning":"consequentialist","policy":"unclear","emotion":"mixed"} ]