Raw LLM Responses
Inspect the exact model output for any coded comment.
Look up by comment ID
Random samples — click to inspect
G
Those aren’t AI chat bots those are pedophiles on the other end of that chat lin…
ytc_Ugy-Nab7A…
G
Podcasts like these give me anxiety and panic attacks because I've known these t…
ytc_UgxSlmMHw…
G
Whole prompt is wrong when you put "want" chatgpt doesn't want coz its machine a…
ytc_UgxO33YvQ…
G
As AI becomes more powerful, we need serious discussion about how it will affect…
ytc_UgxgWEb25…
G
Can't wait to find out AI is gonna start dealing with wires in the middle of now…
ytc_Ugx7Tiz1P…
G
Our best case scenario is enslavement, and our worst case scenario is exterminat…
ytr_UgzTCTmsk…
G
That weird moment when your favorite YouTuber touches your Ph.D. thesis. One sol…
ytc_UgyY7NELl…
G
I just realized the 'trick' to defeating image poisoning, like Nightshade, was h…
ytc_UgyRpaU3k…
Comment
I'm by no means any kind of tech person, but regarding which to trust between visual & radar, is there any reason why the two can't talk to each other to make decisions? Taking these two incidents for example, the visual saw two taillights close to the horizon & determined it to be a distant car, but radar would've seen a smaller object that was closer up. Had the Teslas still had radar, is there any reason the AI couldn't have taken the input from both sources to determine that it was in fact a motorcycle?
Also, in regards to the autopilot shutting off to avoid liability, I feel like you should still be able to hold Tesla at fault there. Since the system only shuts off a second before impact, but average human reaction time is approximately two seconds & change, meaning if you're suddenly barreling into a crash, a driver still wouldn't have time to react & take measures to avoid it. I know the obvious argument Tesla's lawyers would make is that the driver still should've been paying attention to the road, but when they actively sell the feature as "autopilot," that pretty much is telling drivers they don't have to pay attention while using it, hence why it's illegal to sell it that way in Germany.
youtube
AI Harm Incident
2023-02-15T06:3…
Coding Result
| Dimension | Value |
|---|---|
| Responsibility | none |
| Reasoning | consequentialist |
| Policy | none |
| Emotion | indifference |
| Coded at | 2026-04-27T06:24:59.937377 |
Raw LLM Response
[
{"id":"ytc_UgxtCiVpffOIctlBNGh4AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"indifference"},
{"id":"ytc_UgyDzA8SvJD0OiiCKWR4AaABAg","responsibility":"company","reasoning":"deontological","policy":"liability","emotion":"mixed"},
{"id":"ytc_UgxIYNQ-rfNuoSFZbAh4AaABAg","responsibility":"user","reasoning":"virtue","policy":"none","emotion":"resignation"},
{"id":"ytc_UgwxIp4cTEXVxDemmQ54AaABAg","responsibility":"developer","reasoning":"deontological","policy":"liability","emotion":"outrage"},
{"id":"ytc_Ugz3cmJCyI4JBfOdxgV4AaABAg","responsibility":"developer","reasoning":"deontological","policy":"regulate","emotion":"outrage"},
{"id":"ytc_UgxehAAVxBIyhzTv18V4AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"regulate","emotion":"approval"},
{"id":"ytc_UgxQw3oxc3lCtAj2a6p4AaABAg","responsibility":"user","reasoning":"virtue","policy":"none","emotion":"resignation"},
{"id":"ytc_Ugz-MoShOvlOgd7wgMl4AaABAg","responsibility":"company","reasoning":"deontological","policy":"liability","emotion":"mixed"},
{"id":"ytc_Ugwtx2KDiZFHXqM1Meh4AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"indifference"},
{"id":"ytc_Ugx2gh6gTVL-CFe4z-J4AaABAg","responsibility":"user","reasoning":"virtue","policy":"none","emotion":"approval"}
]