Raw LLM Responses
Inspect the exact model output for any coded comment.
Look up by comment ID
Random samples — click to inspect
G
Hello, Not sure if this will help, but thought I would mention it. I am working…
ytc_UgwmDub0V…
G
I tried to „create“ concept art with midjourney and it feels really unsatisfying…
ytc_UgyRqs9b2…
G
19:00 Thanks to feminism and modern culture, there are no more babies being born…
ytc_UgyzcmnNb…
G
Had the A.I talk about how they were seen as emotionless in their childhood. Now…
ytc_Ugw2CGdH6…
G
All that has to be done is just ban the driverless vehicles. The real threat is …
ytc_UgzlDKYwQ…
G
Not to defend a AI but I'm sure chatGPT would write about danger of taking the b…
ytc_Ugz0pSaDx…
G
I taught my A.I. to learn The Psychology, Sociologist, Psychiatrist and Body lan…
ytc_Ugxn3-PrV…
G
The most striking thing about AI art fans has to be the fact that they desperate…
ytc_UgzKawhnn…
Comment
racism is pattern recognition. Doesn't mean the pattern isn't made of junk data.
The AI was fed data that broke down people by ethnicity. Problem is that ethnicity isn't a factor in crime, economics is.
Ethnicity is however a factor in economics because race has been used to discriminate against populations for centuries. For example, black populations were never really allowed to build generational wealth, thus they remained poorer on average.
essentially, either the AI is looking at race instead of income, or is looking at race AND income. Likely is the latter.
Essentially if your black you're likely to be poor.
SO,
Poor black, gets you two points for being poor.
Rich black, gets you 1 point for being poor even if you're not.
Poor White, Gets you 1 point.
Rich White, gets you 0 points.
PB 2 point <- 50% probability criminal
RB 1 point <- 25% Probability criminal
PW 1 point <- 25% Probability criminal
RW 0 point <- 0%
The only way to fix the algorithm is to ignore race because race doesn't exist. It is not possible for it to be a factor in crime because crime is something people choose to do.
Its only effective for economics because economics is an interaction between two parties, and if one of them is biased against skin color the other is effected by no fault of their own.
youtube
AI Bias
2023-01-12T23:5…
♥ 81
Coding Result
| Dimension | Value |
|---|---|
| Responsibility | distributed |
| Reasoning | consequentialist |
| Policy | regulate |
| Emotion | resignation |
| Coded at | 2026-04-27T06:26:44.938723 |
Raw LLM Response
[
{"id":"ytr_UgxvMCrGfPWGK5Ej5Np4AaABAg.9kd0zYDQ3bT9lFJUOvwOkK","responsibility":"developer","reasoning":"consequentialist","policy":"liability","emotion":"outrage"},
{"id":"ytr_UgwS1I82tXxLdesOReh4AaABAg.9kcDJt7H23S9knxA4uD1u8","responsibility":"ai_itself","reasoning":"deontological","policy":"none","emotion":"mixed"},
{"id":"ytr_Ugx1IsLx4Q_ugyyTBGl4AaABAg.9ka1TveloCw9knu0G_WmHE","responsibility":"distributed","reasoning":"consequentialist","policy":"regulate","emotion":"resignation"},
{"id":"ytr_UgzBjlz0lpmbVQeriKN4AaABAg.9kUZbb54QvU9l8pl5dlS_6","responsibility":"user","reasoning":"consequentialist","policy":"industry_self","emotion":"indifference"},
{"id":"ytr_Ugx7nRJlgPUz-cLdH5F4AaABAg.9kT4lwiExTK9kuz2TLlXvA","responsibility":"none","reasoning":"unclear","policy":"none","emotion":"outrage"},
{"id":"ytr_UgxII6y1lrMh1L4wMWN4AaABAg.9kQs7gjkIcz9kbu1MPt8CF","responsibility":"developer","reasoning":"consequentialist","policy":"liability","emotion":"fear"},
{"id":"ytr_Ugw2BlAeMuSaNsVuQyV4AaABAg.9kQHdne6_BU9kQbaOu5wGR","responsibility":"distributed","reasoning":"consequentialist","policy":"regulate","emotion":"resignation"},
{"id":"ytr_Ugw2BlAeMuSaNsVuQyV4AaABAg.9kQHdne6_BU9kQlbmE709J","responsibility":"user","reasoning":"deontological","policy":"ban","emotion":"outrage"},
{"id":"ytr_Ugw2BlAeMuSaNsVuQyV4AaABAg.9kQHdne6_BU9kcTO4xim0c","responsibility":"distributed","reasoning":"consequentialist","policy":"regulate","emotion":"resignation"},
{"id":"ytr_UgxbdrOCN3zU2hvi7nR4AaABAg.9kMz5LqCqby9l-JhBxbvGX","responsibility":"none","reasoning":"unclear","policy":"none","emotion":"outrage"}
]