Raw LLM Responses

Inspect the exact model output for any coded comment.

Look up by comment ID
Random samples — click to inspect
G Can't believe what I'm reading here in the comments. "Her reaction is cringe" … ytc_Ugz_gXql6… G Your phone, washing machine, coffee machine, TV, tablet, PC, laptop, medical equ… ytc_Ugx3yRlhs… G We need a video called "why Neil deGrass Tyson is overrated - With AI". Seriousl… ytc_UgwUR9uCo… G If we went back in time and start AI artists to use these tools in replacement o… ytr_Ugwf5jmSh… G I am predicting that AI based on LLM architecture will never achieve AGI. No amo… ytc_Ugx5m86V0… G ChatGPT is programmed by people, humans. All it's responses are human content le… ytc_UgztMCgo0… G I don’t normally watch YouTube videos in the art sphere. This vid could serve as… ytc_Ugxizq4RL… G The Godfather is mad he isn't making enough money😂 AI can't turn into Skynet. Wh… ytc_UgwuYyX9r…
Comment
This falls under the long tail problem. They can build an AI to handle 99.9% of the common cases, but then there are still a huge number of uncommon cases that they have to deal with almost one by one to get the last 0.1%. They just recently made a big improvement to their handling of unprotected left turns and that was a big effort. Ultimately I believe some form of RADAR will come back once they improve their handling of sensor fusion. The car should be able to have multiple sensors such as cameras, RADAR, ultra sonics, microphones, map data, etc. and be able to blend all those inputs together the way our brains do to make better decisions. But sensor fusion is currently a very difficult problem, so they aren't there yet. In the meantime we need better countermeasures on motorcycles to protect ourselves in these situations, like by having rear facing radar to warn of a car closing on us too fast. To help the AI see us, a rear facing radar could trigger a warning like flashing the brake light.
youtube AI Harm Incident 2022-09-03T15:2… ♥ 3
Coding Result
DimensionValue
Responsibilitycompany
Reasoningconsequentialist
Policyindustry_self
Emotionresignation
Coded at2026-04-27T06:24:59.937377
Raw LLM Response
[ {"id":"ytc_Ugz7sxzninZ8dCjx35R4AaABAg","responsibility":"user","reasoning":"consequentialist","policy":"liability","emotion":"fear"}, {"id":"ytc_Ugz7IpfUAxm6ju_kM_l4AaABAg","responsibility":"company","reasoning":"consequentialist","policy":"regulate","emotion":"outrage"}, {"id":"ytc_UgyNEJadqW4kR263ScF4AaABAg","responsibility":"company","reasoning":"deontological","policy":"ban","emotion":"outrage"}, {"id":"ytc_UgyORXstL9nvS-ytVA14AaABAg","responsibility":"company","reasoning":"mixed","policy":"regulate","emotion":"mixed"}, {"id":"ytc_UgyhkMKVgvWdPApJAgZ4AaABAg","responsibility":"user","reasoning":"virtue","policy":"none","emotion":"indifference"}, {"id":"ytc_UgyiVTXmNo-D3092PWx4AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"industry_self","emotion":"approval"}, {"id":"ytc_Ugw8zgJOBq3qUiFXOxZ4AaABAg","responsibility":"company","reasoning":"deontological","policy":"regulate","emotion":"outrage"}, {"id":"ytc_Ugzuh8j1CSoi8aEOWkl4AaABAg","responsibility":"company","reasoning":"deontological","policy":"regulate","emotion":"outrage"}, {"id":"ytc_UgxabLh50kl1NoJeRrF4AaABAg","responsibility":"company","reasoning":"consequentialist","policy":"industry_self","emotion":"resignation"}, {"id":"ytc_UgzNJVrWN-ubPT-Ab-V4AaABAg","responsibility":"company","reasoning":"deontological","policy":"regulate","emotion":"outrage"} ]