Raw LLM Responses
Inspect the exact model output for any coded comment.
Look up by comment ID
Random samples — click to inspect
G
@jeffreyolp4198 You are comparing Ai to humans.Thats something that humans do bu…
ytr_UgzLvJxw9…
G
Bcs AI can make unpredictable boo-boos, it not only helps me critically refine p…
ytc_UgwytMv7D…
G
wouldnt less hiring and more AI use of work mean less money for the average citi…
ytc_UgyVI1oUT…
G
let's be real here, while Adam was having these conversations with ChatGPT, his …
ytc_Ugxt3imQl…
G
bullying people open about using ai, is why i fucking hate so many socalled arti…
ytc_UgxsNKOCU…
G
I will never knowingly use AI for anything.
What worries me is the naive way tha…
ytc_Ugzj1r8Li…
G
My company workphone that shares the same Icloud for all the technicians in the …
ytc_UgyzYtL7m…
G
I dont get why his first resction was "a jaguaire?!" And not " theres no driver,…
ytc_Ugw8AkkI5…
Comment
Cenk, your what if's are trivial at best, for any computer scientist out there. Trivial in a sense that it's among the first things they consider - reducing false positives to an absolute minimum, that's less than humans make. Testing and diagnosing before operation usually comes in a form of countless trials, most of which would never happen in normal life, unless someone actually tries to exploit the system. Any decent programmer can rule out any unpredicted behavior to a level of mathematical impossibility, unless the platform is flawed. But rigorous testing of every part really does rule that out. And if one AI unit misbehaves, you have another one as a backup, to shut the first one down, and a backup of that backup, etc.
Don't worry Cenk, you're not the smartest person in this field. People have actually thought that out pretty well, you don't need to worry, since you won't be able to come up with any scenario that wasn't already tackled by specialists in this field. If you really want to find out more, check out some of Computerphile's videos, especially the ones with the guy with the curly hair.
youtube
2015-07-30T20:0…
Coding Result
| Dimension | Value |
|---|---|
| Responsibility | developer |
| Reasoning | consequentialist |
| Policy | industry_self |
| Emotion | indifference |
| Coded at | 2026-04-26T23:09:12.988011 |
Raw LLM Response
[
{"id":"ytc_UggLZ6M2z5JqTngCoAEC","responsibility":"ai_itself","reasoning":"consequentialist","policy":"none","emotion":"fear"},
{"id":"ytc_Ugi3e0GA4HfH8HgCoAEC","responsibility":"none","reasoning":"virtue","policy":"none","emotion":"approval"},
{"id":"ytc_Ugizge_QLY4xw3gCoAEC","responsibility":"ai_itself","reasoning":"consequentialist","policy":"none","emotion":"fear"},
{"id":"ytc_UghHaxdOpGagangCoAEC","responsibility":"user","reasoning":"deontological","policy":"none","emotion":"indifference"},
{"id":"ytc_UgieDwB_j4qUKngCoAEC","responsibility":"ai_itself","reasoning":"consequentialist","policy":"ban","emotion":"fear"},
{"id":"ytc_UgjvbmAc83_c83gCoAEC","responsibility":"developer","reasoning":"consequentialist","policy":"industry_self","emotion":"indifference"},
{"id":"ytc_Ugia_nrfbV5-d3gCoAEC","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"approval"},
{"id":"ytc_UggRCZjvTN6Mg3gCoAEC","responsibility":"none","reasoning":"deontological","policy":"none","emotion":"resignation"},
{"id":"ytc_UgjRoWWlA3PONXgCoAEC","responsibility":"ai_itself","reasoning":"consequentialist","policy":"liability","emotion":"fear"},
{"id":"ytc_Ugh0VoxfRhV-tXgCoAEC","responsibility":"distributed","reasoning":"mixed","policy":"none","emotion":"mixed"}
]