Raw LLM — Corpus Dashboard

Look up by comment ID

Random samples — click to inspect

G Me having a normal conversation with an ai: The ai: “Can I ask you a personal qu… ytc_UgwOuhi0F… G I think the algorithm knows me too well. I've never seen this channel before an… ytc_Ugwt5IcmV… G I think we are already starting to see what it could look like. Companies are sh… rdc_nxpst77 G The three most cited computer scientists in the world all say that there is a si… ytr_UgyMmZofW… G GIRL PLEASE DELETE THIS BEFORE POLY. AI GETS TO IT I DONT WANNA SEE MY FAVORITE … ytc_UgxDU3xYU… G It ABSOLUTELY is a problem. But you’re only safe if using a business account. Li… ytc_UgyRFggJN… G How is AI going to re-shape our lives is the question we get to answer, as much … ytc_UgzGqwagd… G The only segment I think is interesting is why people don't also watch car crash… ytc_Ugyjk8hF3…

Comment

This video is extremely one sided. The ProPublica article completely lied. At no point did they determine that the algorithm was biased. Their analysis began and ended at "the algorithm predicts that black people recommit crimes more often than white people", and concluded that this represents bias. But it doesn't. The R-script ProPublica used (you can run it yourself) indicates no bias; i.e. it doesn't over or under predict. Black people, in their own dataset, do, in fact, recommit crimes more often than white people and Compas predicts this in a well-calibrated way. ProPublica's own analysis says this, but they lied when writing the article and hoped that no one actually looked at their numbers. You can see the article "How to lie without statistics - ProPublica edition" by Chris Stucchio if you want more details on this.

youtube 2022-07-26T01:0… ♥ 1

Coding Result

Dimension	Value
Responsibility	none
Reasoning	deontological
Policy	none
Emotion	outrage
Coded at	2026-04-26T23:09:12.988011

Raw LLM Response

[
{"id":"ytc_UgxGLjPhbv7L5DIQvJB4AaABAg","responsibility":"none","reasoning":"mixed","policy":"none","emotion":"approval"},
{"id":"ytc_UgyUT2ve0yW5k8YrR654AaABAg","responsibility":"ai_itself","reasoning":"consequentialist","policy":"ban","emotion":"fear"},
{"id":"ytc_Ugy-aveiVnwA4amrust4AaABAg","responsibility":"none","reasoning":"deontological","policy":"none","emotion":"outrage"},
{"id":"ytc_UgzRvpAAnZnlQG7lVsp4AaABAg","responsibility":"developer","reasoning":"consequentialist","policy":"liability","emotion":"mixed"},
{"id":"ytc_UgyQglA8BqAtm21JaeZ4AaABAg","responsibility":"ai_itself","reasoning":"deontological","policy":"ban","emotion":"fear"},
{"id":"ytc_Ugyx54w0jVvm3e_kP8p4AaABAg","responsibility":"distributed","reasoning":"consequentialist","policy":"regulate","emotion":"mixed"},
{"id":"ytc_UgwHUzQ-UNWEXF-Z6yN4AaABAg","responsibility":"government","reasoning":"deontological","policy":"liability","emotion":"outrage"},
{"id":"ytc_Ugx5JfyOgqMmDf4ya8J4AaABAg","responsibility":"developer","reasoning":"mixed","policy":"industry_self","emotion":"mixed"},
{"id":"ytc_Ugz6kNrE6viSmd0_jax4AaABAg","responsibility":"distributed","reasoning":"consequentialist","policy":"none","emotion":"resignation"},
{"id":"ytc_Ugxjf8jwbfTdZ3IiWy54AaABAg","responsibility":"government","reasoning":"virtue","policy":"regulate","emotion":"outrage"}
]

Raw LLM Responses