Raw LLM — Corpus Dashboard

Look up by comment ID

Random samples — click to inspect

G Im only 16 so i know ive got plently of time to chose something else. But art is… ytc_UgxQHVGyy… G Maybe if consumers actually raised their standards, "slop" wouldn't be so entici… ytr_Ugwph_3TM… G We are starting a dangerous new chapter for this internet/media world. Before i… rdc_kjn771c G Why did anyone think self-driving cars was a good idea? I mean come on, this is … ytc_UgxJ9MaVc… G If it takes away from jobs, then those that think they themselves will profit fr… ytc_Ugz69yb-b… G But we've been building the global economy together our whole lives. And now she… ytc_UgwWMIyKC… G Nice that one of your arguments for hotdogs being sandwiches is the same as AI's… ytc_UgxrSiwTM… G imagine you cornered chatGPT like alex did and it just starts spitting out your … ytc_UgxuCRrQQ…

Comment

This really isn’t going to work long term. The whole reason AI datasets are so large is to fix any weird holes anyway. A lot of AI training data is already poisoned by SEO keywords and algorithmic BS. Looking at the Nightshade article, they had to train a Lora on 300 poisoned images to really make a massive difference - but that’s a dataset of ONLY poisoned images. If there’s 10,000 poisoned images in a dataset of 50 million, it’s a statistical anomaly. Artists would better spend their time working with companies like Invoke AI who want to help safeguard artist rights, and find ways to make say opt-out forms and such to have their dara excluded. Some AI companies already allow artists to opt out - Stable Labs removed Greg whatshisname from their dataset at his request. Yes, you’ll never be able to prevent unscrupulous people from training Lora sets or something. But software companies have likewise had to learn that DRM never fully prevents unscrupulous people from pirating software either.

youtube Viral AI Reaction 2024-11-04T14:1…

Coding Result

Dimension	Value
Responsibility	ai_itself
Reasoning	consequentialist
Policy	none
Emotion	resignation
Coded at	2026-04-27T06:24:59.937377

Raw LLM Response

[
{"id":"ytc_UgyMd8RbSEnUXJhChFh4AaABAg","responsibility":"government","reasoning":"consequentialist","policy":"regulate","emotion":"outrage"},
{"id":"ytc_UgxbEBMYtTR1dnzrJJd4AaABAg","responsibility":"user","reasoning":"virtue","policy":"none","emotion":"disapproval"},
{"id":"ytc_UgxLwMI2gLsQyF5FqjR4AaABAg","responsibility":"ai_itself","reasoning":"consequentialist","policy":"none","emotion":"fear"},
{"id":"ytc_UgwkAOtFoAVOVNvzLd14AaABAg","responsibility":"user","reasoning":"virtue","policy":"none","emotion":"disapproval"},
{"id":"ytc_Ugz4_mwpA8tmAylj-Ax4AaABAg","responsibility":"company","reasoning":"deontological","policy":"regulate","emotion":"outrage"},
{"id":"ytc_Ugx3NmgZO6I6xwAfJi14AaABAg","responsibility":"company","reasoning":"contractualist","policy":"regulate","emotion":"mixed"},
{"id":"ytc_Ugyu-l4SGUvXOyQjZIN4AaABAg","responsibility":"ai_itself","reasoning":"consequentialist","policy":"none","emotion":"resignation"},
{"id":"ytc_UgziofpHglQxBhnrOR14AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"approval"},
{"id":"ytc_Ugw28z27FcR-uTmP2x94AaABAg","responsibility":"distributed","reasoning":"mixed","policy":"none","emotion":"mixed"},
{"id":"ytc_UgxGcaMW7bB-1bDbhSx4AaABAg","responsibility":"none","reasoning":"deontological","policy":"none","emotion":"indifference"}
]

Raw LLM Responses