Raw LLM Responses
Inspect the exact model output for any coded comment.
Look up by comment ID
Random samples — click to inspect
G
It’s funny because the last one tries to make Gemini (formerly Bard) look worse …
ytc_Ugz23wW9i…
G
This needs a signature campaign to get it answered by public vote. I, for one, d…
ytc_Ugx7bSZDm…
G
> Harari defines AI as “Alien Intelligence”, implying that we no longer know …
rdc_oerozxk
G
AI can actually be really cool unlock really good I bet if I showed you one of t…
ytc_Ugx5jYHds…
G
Lol fucking nerd.
How bout not giving them access to that shit? Like wtf are we …
ytc_Ugy5dYgEw…
G
Others - ahhh i got mini heart attack in last clip
Legends -chil bro it is ai g…
ytc_UgwqW7T7g…
G
Problem is there’s no such thing as Ai it’s not intelligent it’s programmed by s…
ytc_UgyBQIM9U…
G
If you understand the technology then you know that the jobs being wiped out are…
ytc_UgwUS9iMS…
Comment
[More.](https://www.techdirt.com/articles/20160623/06112634789/proprietary-algorithms-are-being-used-to-enhance-criminal-sentences-preventing-defendants-challenging-them.shtml)
The algorithms are made by someone (a 'private company' in this case), and it determines how long the sentence should be. No-one is allowed to check how this algorithm works, therefore it is as reliable as some random dude making a claim about the length of a sentence.
Here's why:
* The article mentions that the punishment depends on things like age and gender, which is against the fundamentals of the law. The law states that everyone should be punished equally for equal deeds.
* If the designer of the algorithm commits a crime, (s)he'll probably get a very low sentence, because (s)he decides the punishment.
* If I have offended that 'private company' in some way, I can expect a longer sentence, because that 'private company' can have things like that taken into their algorithm.
People should understand that computers aren't magical machines. Computers are build by people, and they do what the designers want the computer to do. And if the designer wants the computer to look like it's playing fair, while secretly cheating, then the computer will do that. You cannot trust any computer if you do not trust the manufacturer.
reddit
AI Harm Incident
1467011205.0
♥ 7
Coding Result
| Dimension | Value |
|---|---|
| Responsibility | company |
| Reasoning | deontological |
| Policy | regulate |
| Emotion | outrage |
| Coded at | 2026-04-25T08:33:43.502452 |
Raw LLM Response
[
{"id":"rdc_oi3xp5l","responsibility":"none","reasoning":"unclear","policy":"unclear","emotion":"indifference"},
{"id":"rdc_d4pl1m1","responsibility":"company","reasoning":"deontological","policy":"regulate","emotion":"outrage"},
{"id":"rdc_jtqvrwv","responsibility":"none","reasoning":"unclear","policy":"unclear","emotion":"approval"},
{"id":"rdc_jts1or8","responsibility":"none","reasoning":"unclear","policy":"unclear","emotion":"approval"},
{"id":"rdc_jtrqmmm","responsibility":"ai_itself","reasoning":"mixed","policy":"unclear","emotion":"mixed"}
]