Raw LLM Responses
Inspect the exact model output for any coded comment.
Look up by comment ID
Random samples — click to inspect
G
Rob Miles has got a great channel on AI safety. And a bunch of videos on Compute…
ytc_Ugz8uz5IU…
G
My only gripe is that ai has no means of reproduction in the way biological bein…
ytc_UgwK23KOP…
G
It's funny how we are saying that we will reach this AI or that AI in X number o…
ytc_UgykiNzjO…
G
Hollywood is dead and not because of AI .. some people won’t be complicit with t…
ytc_UgzY0YlO5…
G
8:23 using ai for therapy is as effective as YOU are.
Essentially, at its most …
ytc_UgzDt8NEw…
G
The idea of using ChatGPT to fool AI detectors is intriguing! I've been explorin…
ytc_UgxXTirVq…
G
The day that AI can write code and program and build other machines is when huma…
ytc_Ugyf-MRj4…
G
Nah but if it was just a null value issue, I feel like Claude would find that pr…
rdc_oi09evs
Comment
Everyone is concerned with these companies selling or sharing their data inappropriately.
I think the *real* issue is that they likely won't share their data *about you* ... *with you*.
At least in my experience *that* seems to be how both the existing medical system *and* this system of the "future" totally fucks up HIPAA. Failing to properly communicate a diagnosis.. failing to provide my *complete* medical history and *all* of the data they have on me is something that happens *frequently*.
If you're marked as a "frequent flyer" or "drug seeker" in a medical system, do you think they share that with you? They pull data on you from various state and federal databases when they prescribe certain drugs now.. which are basically "black boxes" ... do they share *any* of that? That they even used one of these databases or "methods" to determine your treatment?
*fuck no*.
Seems like it might get a lot worse with folks like google, as the data analysis ... and stuff like machine learning will complicate the nature of the data they have on you and *how* they can even communicate it all to you.. etc. So now not only will they refuse to communicate stuff that you might not "like", they'll fail to communicate stuff that is essential to your health...
... and not necessarily because they can't, but because it's "hard".
reddit
AI Bias
1574930392.0
♥ 1
Coding Result
| Dimension | Value |
|---|---|
| Responsibility | company |
| Reasoning | deontological |
| Policy | liability |
| Emotion | outrage |
| Coded at | 2026-04-25T08:33:43.502452 |
Raw LLM Response
[
{"id":"rdc_nk48jjs","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"resignation"},
{"id":"rdc_f8xreyu","responsibility":"company","reasoning":"deontological","policy":"liability","emotion":"outrage"},
{"id":"rdc_erpg6xh","responsibility":"company","reasoning":"virtue","policy":"unclear","emotion":"outrage"},
{"id":"rdc_ewm3x6p","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"outrage"},
{"id":"rdc_ewmk9tz","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"approval"}
]