Raw LLM Responses

Inspect the exact model output for any coded comment.

Comment
It seems the only consideration the programmers have provided to simulate depth perception is "If single tail light shut off autopilot" or "lateral spacing of tail lights determines distance: Near = wide spacing/far = close spacing." Unfortunately a close motorcycle looks like a car at a safe distance. A lucid human would be able to discern other, subtle cues such as: 1. Comparing the size of familiar highway terrain features to the size of any vehicle (medians, barriers, fenceposts, signs, reflectors, surface paint stripes etc). These features are so varied it is impossible to reasonably predict with hard data. Sufficient amounts of data collected would be immense, exceeding storage capacity and timely, useful access. Is there a possibility of broadening existing hardware/software to include or increase not only new data, but also sensitivity modalities of differing intensities and "emergency mode" data access (like adding adrenaline)? 2. The motorcycle's headlight casts light ahead of it but also provides for "backlighting" as viewed from the rear. This creates a discernable "shadow silhouette" evident to an approaching (human) driver. This effect is like looking up for the presence of stars and seeing dark spaces void of light and knowing there are either clouds or mountains obscuring them from view. The light pollution of a highway environment frustrates a human's ability to see this at night, but could an array of cameras be programmed to notice this? 3. A human would be aware that motorcycles tilt to turn, their suspensions move more rapidly vertically, and feature other motions peculiar enough to arouse suspicion of close proximity, not to mention their common acoustic signatures. A young person new to driving would not be looking for these things as much as a seasoned one, yet they as subjective humans would still be much more likely to brake than a computer with such narrow programming parameters. AI and its programmers are going to kill, imprison, injure and destroy more innocent people during the quest to create intelligence they hope will exceed human capacity. They're not even close at present. **You can't standardize light spacing without violating freedom of motorcycle design and customization. You could design and provide motorcycle transponders, but not all people would know about or use them. Or...you could just go back to radar.
youtube AI Harm Incident 2022-09-03T18:2…
Coding Result
DimensionValue
Responsibilitydeveloper
Reasoningdeontological
Policyunclear
Emotionfear
Coded at2026-04-27T06:24:59.937377
Raw LLM Response
[ {"id":"ytc_UgxchwfMLlIwrAJI3op4AaABAg","responsibility":"company","reasoning":"consequentialist","policy":"none","emotion":"indifference"}, {"id":"ytc_Ugyl8DnJWm-IYVOcjY54AaABAg","responsibility":"developer","reasoning":"deontological","policy":"unclear","emotion":"fear"}, {"id":"ytc_Ugy2N48LIE3brL-Jq294AaABAg","responsibility":"company","reasoning":"virtue","policy":"ban","emotion":"outrage"}, {"id":"ytc_UgxUK0-sdyWUaZ_KXnl4AaABAg","responsibility":"user","reasoning":"consequentialist","policy":"industry_self","emotion":"approval"}, {"id":"ytc_Ugzi39MZlu6PbG-zePR4AaABAg","responsibility":"user","reasoning":"deontological","policy":"none","emotion":"fear"}, {"id":"ytc_UgxFvXkriaARSpnDKJp4AaABAg","responsibility":"developer","reasoning":"deontological","policy":"regulate","emotion":"outrage"}, {"id":"ytc_UgwPAJuRYVc0pj9scnV4AaABAg","responsibility":"none","reasoning":"unclear","policy":"unclear","emotion":"indifference"}, {"id":"ytc_UgwfJziH3uN0rzxQCrl4AaABAg","responsibility":"government","reasoning":"consequentialist","policy":"regulate","emotion":"approval"}, {"id":"ytc_UgxS5e_Af-V7IXSS2Q94AaABAg","responsibility":"company","reasoning":"consequentialist","policy":"industry_self","emotion":"approval"}, {"id":"ytc_UgxnaADHhErV5QLsoDV4AaABAg","responsibility":"company","reasoning":"deontological","policy":"liability","emotion":"fear"} ]