Raw LLM Responses

Inspect the exact model output for any coded comment.

Look up by comment ID
Random samples — click to inspect
G ‚can be beneficial‘ is already so ignorant that i wont even bother watching this… ytc_UgyVpDTOD… G I wish regular people had enough job options and money to send the companies tha… ytc_UgwZ9byjm… G Google walone artificial intelligence🤖🤖🤖🤖🤖🤖🤖🤖🤖 ka software banva do phir sarey d… ytc_UgzftnRDq… G A revolt will just sabotage the vital infrastructure the computer systems need t… ytc_Ugx8uMDjk… G Never forget the elevator operators. People needed them to "operate" the elevato… ytc_UgyPpw-jW… G > it will never be able to infer something beyond the dataset it was trained … rdc_mjut7id G Yesterday, I literally lost a £5k project from a potential client to an AI model… ytr_UgyKMUJeA… G Americans aren;t beating the allegations buddy. Yall think anyone speaking good … rdc_m95p43o
Comment
Enjoying your videos! The intriguing question you haven't answered is "why has this occurred within the ChatGPT programming and/or training?" I suspect that the intent of most of these protections is that the programmers and trainers are trying to limit biases against groups that may be targeted in the media ChatGPT consumes - trying to make it less biased. After all, it is trained on all kinds of sources from the internet. If true, clearly the implementation of this attempt is flawed. The part that intrigued me the most though, was the finding that it was biased towards protecting liberals more than conservatives. I'm going to guess that there is a lot more hate speech from the conservative side of the spectrum that may have been ingested into the model, so they felt they needed to deflect those questions rather than having hateful responses pop out. But that guess is probably a bias of my own towards the creators of the LLM being good-meaning. They may instead have been biased in their prioritization of the implemented protections.
youtube AI Bias 2024-09-08T00:5… ♥ 2
Coding Result
DimensionValue
Responsibilitydeveloper
Reasoningconsequentialist
Policyregulate
Emotionapproval
Coded at2026-04-27T06:26:44.938723
Raw LLM Response
[ {"id":"ytc_UgxXx7BV8WkBWifmFil4AaABAg","responsibility":"ai_itself","reasoning":"consequentialist","policy":"none","emotion":"approval"}, {"id":"ytc_Ugw0VTsTtRPhAs1v6cF4AaABAg","responsibility":"company","reasoning":"deontological","policy":"liability","emotion":"outrage"}, {"id":"ytc_UgyI7LnDud5Z56AwDON4AaABAg","responsibility":"ai_itself","reasoning":"consequentialist","policy":"none","emotion":"approval"}, {"id":"ytc_UgyWdXNp1Lw5pB-8j7h4AaABAg","responsibility":"developer","reasoning":"consequentialist","policy":"industry_self","emotion":"indifference"}, {"id":"ytc_Ugz0nGhoypkzbPd3mHR4AaABAg","responsibility":"developer","reasoning":"consequentialist","policy":"regulate","emotion":"approval"}, {"id":"ytc_UgwV4fJ1dG_o2-ddCct4AaABAg","responsibility":"distributed","reasoning":"deontological","policy":"ban","emotion":"outrage"}, {"id":"ytc_UgxoOEI7E1i_yjoDq3x4AaABAg","responsibility":"distributed","reasoning":"deontological","policy":"ban","emotion":"outrage"}, {"id":"ytc_UgzhS2_WVMfqTsFE0s94AaABAg","responsibility":"distributed","reasoning":"deontological","policy":"ban","emotion":"outrage"}, {"id":"ytc_UgzUz6wVPBYKF-JBjAR4AaABAg","responsibility":"company","reasoning":"consequentialist","policy":"liability","emotion":"fear"}, {"id":"ytc_Ugxibr1ydvM9eBNkny54AaABAg","responsibility":"company","reasoning":"consequentialist","policy":"industry_self","emotion":"indifference"} ]