Raw LLM Responses
Inspect the exact model output for any coded comment.
Look up by comment ID
Random samples — click to inspect
G
i reallly appreciate this video. though i am a super hard hater of AI, lol... y…
ytc_Ugyf2FMcd…
G
I described this short to chatgpt and he said ,*do you think i would make a good…
ytc_UgxUDeilR…
G
kinda weird to be on both sides of this lol, on the one hand expressing your ide…
ytc_UgwA2ERhN…
G
not saying this is or isnt my case, but since I lost my job back in late decembe…
ytc_Ugw8bRNy8…
G
having an ai predict who will commit crimes and adding them to a "heat least", …
ytc_Ugyhn01wQ…
G
@gehteuchnixan69 Dude, your 10 year old VW Golf was much much better at self-dri…
ytr_Ugy2pW-Y1…
G
The solution is to dump social media. Once the ad revenue and view numbers stop …
ytc_UgzG-1aLr…
G
I asked a couple of chatbots about this. Iirc, chatgpt didn't care about please …
ytc_UgzMgBr-6…
Comment
It is unfortunately a cat-and-mouse game. Every time you come up with a solution that prevents Image scraping and training, a AI engineer is working on ways to overcome just that.
Poisoning works because of how it manipulates what the AI sees when if tries to learn from the image.
Edge detection fails when it sees a gargled mess, hence the poisoning works.
BUT... what if the AI does something to the image that normalises the image before processing it? Like, you cant see the poisoning, meaning a human isn't sensitive enough to see the artifacts. What if the bot "fixes" the image to where is doesn't have that artifacting before being fed into the model?
But i agree with the sentiment. If its used without permission, then it is stolen work.
And don't even get me started on using the generated images to harass people. Thats just evil.
I think a paradigm shift is needed and poisoning isn't really the answer. It works... for now... so we use that while the next solution is cooking.
youtube
Viral AI Reaction
2024-10-23T11:4…
Coding Result
| Dimension | Value |
|---|---|
| Responsibility | distributed |
| Reasoning | consequentialist |
| Policy | industry_self |
| Emotion | resignation |
| Coded at | 2026-04-27T06:24:59.937377 |
Raw LLM Response
[
{"id":"ytc_UgyW9vFRi93R8p4YYaF4AaABAg","responsibility":"company","reasoning":"deontological","policy":"regulate","emotion":"outrage"},
{"id":"ytc_Ugx0DIiuMV6CDgwC3xt4AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"indifference"},
{"id":"ytc_UgyHzybdo8CBkQJp-EF4AaABAg","responsibility":"none","reasoning":"virtue","policy":"none","emotion":"approval"},
{"id":"ytc_UgzxByI1M0ovhgcpjYB4AaABAg","responsibility":"distributed","reasoning":"consequentialist","policy":"liability","emotion":"fear"},
{"id":"ytc_UgzGPEsJJl9QbrNArp14AaABAg","responsibility":"none","reasoning":"deontological","policy":"none","emotion":"approval"},
{"id":"ytc_Ugx_4rbncV3qMCh8wKV4AaABAg","responsibility":"ai_itself","reasoning":"mixed","policy":"ban","emotion":"outrage"},
{"id":"ytc_UgxBw3yh5lBxrkesYvV4AaABAg","responsibility":"distributed","reasoning":"consequentialist","policy":"industry_self","emotion":"resignation"},
{"id":"ytc_UgxMYBrZ3ehiZZE02nd4AaABAg","responsibility":"company","reasoning":"consequentialist","policy":"ban","emotion":"outrage"},
{"id":"ytc_UgxOtUGOTiL83en-mOh4AaABAg","responsibility":"none","reasoning":"unclear","policy":"none","emotion":"mixed"},
{"id":"ytc_UgwOqvSQyElZasbXbk54AaABAg","responsibility":"distributed","reasoning":"mixed","policy":"regulate","emotion":"approval"}
]