Raw LLM Responses
Inspect the exact model output for any coded comment.
Look up by comment ID
Random samples — click to inspect
G
There is literally no reason AI won't be able to do all those things in the futu…
ytc_UgwK-9B2p…
G
AI is such a scary but amazing thing to talk about.
The part that scares me is t…
ytc_UgxlymtF7…
G
This personal task robot ,female typewill not be talking back. She she telks; st…
ytr_UgxW9RLea…
G
Perhaps this issue isn't "AI can teach you to make bioweapons", and the real iss…
ytc_Ugx-eLpEr…
G
I am working on something called the Relational Heuristic Framework... I'm hopin…
ytc_Ugy0KYs9J…
G
1: they don't OWN anything, let alone the the sentence they typed into a compute…
ytc_UgzXCChlb…
G
Well, Matt Yglesias says the Bangladeshis should be happy to work in collapsing …
rdc_d3sgsoh
G
AI should help us to work only couple hours everyday and earn the SAME salary. N…
ytc_UgxmaiOE_…
Comment
11:40 And here we are leaving the realm of science and entering the realm of wildly asserting things. Why? Well we waste no time jumping to the conclusion that "race" is a cause of arrest and conviction and not a correlating factor of which we need to know why it exists.
We spend all our time looking at the false positive and false negative defiation of a bail algorithm, but we aren't shown what the respective crime rates are or what share of all crimes were actually commited by the respective groups, simply because it is imputed that people are being falsely convicted while belonging to one group and wrongly acquitted while belonging to a different group.
The "bail or not bail" algorithm accuracy doesn't reflect that context at all. You can NOT infere from a false positive bail denial that convictions are also illegitimate!
It has no bearing on either the rate of arrest or falsely convicted people of certain demographies. Hence the allegation that your crimal record is based on a race bias is an unproven imputation.
We don't know how the arrests of one group relate to the crime rate of said group. We could find out that statistically spoken "not enough" suspects from a certain demographic group were arrested for a certain type of crime. And what would we do then, if that were the case? Would the COPs be supposed to round up some vagrants to make the statistics even?
The judicial system was not given an opportunity to prove its impartiality in this presenation. NO effort was made to prove that certain arrests and convictions were illegimate.
youtube
AI Harm Incident
2020-01-11T23:0…
Coding Result
| Dimension | Value |
|---|---|
| Responsibility | company |
| Reasoning | mixed |
| Policy | none |
| Emotion | mixed |
| Coded at | 2026-04-27T06:24:53.388235 |
Raw LLM Response
[
{"id":"ytc_UgzvhwFvB7oi-_lJH7p4AaABAg","responsibility":"government","reasoning":"consequentialist","policy":"liability","emotion":"fear"},
{"id":"ytc_UgxRsW6_dK9mZHrDGgF4AaABAg","responsibility":"government","reasoning":"deontological","policy":"ban","emotion":"outrage"},
{"id":"ytc_Ugy2g28mjkxLDtGgBq54AaABAg","responsibility":"ai_itself","reasoning":"mixed","policy":"none","emotion":"mixed"},
{"id":"ytc_UgyWo6ork0Nq7swGDrB4AaABAg","responsibility":"user","reasoning":"consequentialist","policy":"industry_self","emotion":"indifference"},
{"id":"ytc_UgzqYe3stj_MCdc8Ont4AaABAg","responsibility":"government","reasoning":"deontological","policy":"regulate","emotion":"outrage"},
{"id":"ytc_Ugwcuweh4_Ouew_cU3t4AaABAg","responsibility":"government","reasoning":"deontological","policy":"ban","emotion":"approval"},
{"id":"ytc_UgxplwNJPzzXQK_DpcJ4AaABAg","responsibility":"company","reasoning":"consequentialist","policy":"liability","emotion":"outrage"},
{"id":"ytc_UgxVnjAzQGbX110Qe0F4AaABAg","responsibility":"company","reasoning":"mixed","policy":"none","emotion":"mixed"},
{"id":"ytc_Ugw9kDZTVOXQyiyQA_l4AaABAg","responsibility":"company","reasoning":"consequentialist","policy":"regulate","emotion":"approval"},
{"id":"ytc_UgyocqvY_rib37g6ROd4AaABAg","responsibility":"ai_itself","reasoning":"mixed","policy":"none","emotion":"mixed"}
]