Raw LLM Responses

Inspect the exact model output for any coded comment.

Look up by comment ID
Random samples — click to inspect
G When ai is weak they have nice words When ai gets stronger they're words will … ytc_UgyDlgXwv… G As a current software engineer AI currently is just a fancy search engine. The … ytc_UgxcIplLB… G I think that bill gates idea to tax automation and AI agent's is genius. It coul… ytc_UgyuaFo3r… G In the end we Will discover that consciousness was a physical property of all ma… rdc_icgltrw G These data centers don't need fresh water. It is simply easier and more cost e… ytc_Ugy3baojI… G Narrow AI seems like a great idea. Assisting human function is the point. We d… ytc_UgxlykFN9… G I think someone who is using a radar cruise control from any GM, Chevy, Honda, T… ytr_UgyIzo9MJ… G Manual labour cannot be automated yet you still need people in the hospitality s… ytc_UgySAuJ7y…
Comment
No, but algorithms can be. These image recognition algorithms, like the convolutional neural network or Mask R-CNN, rely on training data to understand how to detect say faces and different features. If we have a dataset that isn't trained on features of other races, like Joy said, the algorithm will not properly detect the face. This is known as overfitting your model. These algorithms can be corrected with proper datasets and will more accurately detect faces of different races. This is what she's saying. Not that physics or mathematics is biased, but rather these systems we code can unintentionally reflect bias we program into it if we fail to consider all possible use cases. This has always been true in computer science. Like making a program that can multiply numbers but that fails to recognize negative numbers. Math isn't biased against negative numbers but your program might be. To fix the "coded gaze" as she says is to be vigilant and ensure we write code that is going to work efficiently and accurately in all possible use cases.
youtube 2019-12-09T18:3… ♥ 3
Coding Result
DimensionValue
Responsibilitydeveloper
Reasoningdeontological
Policyregulate
Emotionapproval
Coded at2026-04-27T06:26:44.938723
Raw LLM Response
[ {"id":"ytr_UgyCC014PwKw1gNsT0p4AaABAg.90hfpqiVk349WJA4vy5TIB","responsibility":"ai_itself","reasoning":"consequentialist","policy":"none","emotion":"indifference"}, {"id":"ytr_Ugxq7EYgHjzqCwzr5QN4AaABAg.8zH3t2SzQVX9K4CZIyLg9h","responsibility":"developer","reasoning":"deontological","policy":"regulate","emotion":"outrage"}, {"id":"ytr_Ugxq7EYgHjzqCwzr5QN4AaABAg.8zH3t2SzQVX9K4L_hdqvYW","responsibility":"distributed","reasoning":"consequentialist","policy":"industry_self","emotion":"approval"}, {"id":"ytr_UgzOw2_UuId_9WOnrFR4AaABAg.8lVjE3jHhDW9WJAIMN8t2D","responsibility":"ai_itself","reasoning":"consequentialist","policy":"none","emotion":"indifference"}, {"id":"ytr_UgwHeHZdwARg_gZKfkp4AaABAg.8YJ_h7HTSut9KPYEygx24a","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"indifference"}, {"id":"ytr_UgglI8_oCeQpHngCoAEC.8SYNeQx5afP92KfqN5z42b","responsibility":"developer","reasoning":"deontological","policy":"regulate","emotion":"approval"}, {"id":"ytr_Ugi6DHYKx8YUMngCoAEC.8RehwotTueh8fh_z-XPGiY","responsibility":"distributed","reasoning":"consequentialist","policy":"industry_self","emotion":"approval"}, {"id":"ytr_UgjF2qj63lIzHHgCoAEC.8ROnyH1SVdNABw8Ocu5aE8","responsibility":"developer","reasoning":"deontological","policy":"ban","emotion":"outrage"}, {"id":"ytr_UggQG6eUAXHh13gCoAEC.8QvIvb-6MgB92KhcvD9Bg_","responsibility":"developer","reasoning":"consequentialist","policy":"regulate","emotion":"approval"}, {"id":"ytr_UgiOUvYyrawYc3gCoAEC.8QjWWcdzynm92KgtJm6ohm","responsibility":"developer","reasoning":"deontological","policy":"regulate","emotion":"approval"} ]