Raw LLM Responses
Inspect the exact model output for any coded comment.
Look up by comment ID
Random samples — click to inspect
G
Hell Sabine ! I prompted chatgpt to answer to today's transcript : here is his c…
ytc_UgwsMTNkD…
G
It seems like many people seem to be uneducated about AI in general here.
Or hav…
ytr_UgxGdhzOY…
G
I think what he says is right….not just the choices of those who do the technolo…
ytc_UgyjQioO1…
G
- When a person feels, and expresses guilt about what they do, they inwardly kno…
ytc_UgxdOhiOn…
G
I agree. After trying out 5 different therapists, “talking” to ChatGPT is the cl…
rdc_njhj1yj
G
А МНЕ ИНОГДА ЖАЛКО ТАКИХ РОБОТОВ СНЯЛИ ЛИЦО!/А У РОБОТА МЫСЛИ ОО-БОЖЕ ЗАЧЕМ ТЫ Э…
ytc_UgyTr66jY…
G
Even if you live 500 yrs, what AI didn’t get the brain right and you have dement…
ytc_UgxJr3s_X…
G
It is difficult to believe that the wealthiest person in the world, who many con…
ytc_UgyvgXRnX…
Comment
The difference between googling something and asking a chatbot is that, when googling, there's a higher chance of stumbling across information you don't already know that could make you reconsider.
If he was googling "what to replace chlorine with" and found a bunch of articles on sodium bromide, then he might have googled "sodium bromide food" or "sodium bromide recipes" or "sodium bromide taste good?" and then found his way to websites talking specifically about the consequences of ingesting sodium bromide.
Of course, it's not guaranteed, and he could have ignored those pages. But getting a variety of information sources presented to you independently, without the chatbot's built in deference or the attempt to maintain a context window across what are in reality different topics, means you're more likely to stumble upon some story of some poor sod who'd previously gone down the same rabbit hole and has already experienced the consequences for you to learn from. Not gonna find that article, that video, that Reddit anecdote, if you're relying only on one source of information, the chatbot, for all your information!
youtube
AI Harm Incident
2026-01-09T07:2…
Coding Result
| Dimension | Value |
|---|---|
| Responsibility | none |
| Reasoning | consequentialist |
| Policy | none |
| Emotion | indifference |
| Coded at | 2026-04-27T06:24:53.388235 |
Raw LLM Response
[
{"id":"ytc_Ugy-RekoDvgI-rCbfeh4AaABAg","responsibility":"developer","reasoning":"deontological","policy":"liability","emotion":"approval"},
{"id":"ytc_UgwmLBZOytjav2-DwCt4AaABAg","responsibility":"user","reasoning":"consequentialist","policy":"none","emotion":"indifference"},
{"id":"ytc_Ugz_PJSxm_4PCc3JWMN4AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"indifference"},
{"id":"ytc_UgzhRHOI1SB1byVJnQt4AaABAg","responsibility":"none","reasoning":"unclear","policy":"none","emotion":"indifference"},
{"id":"ytc_UgzGL_JXmVBhqqcmFNJ4AaABAg","responsibility":"none","reasoning":"mixed","policy":"none","emotion":"approval"},
{"id":"ytc_UgzN_NlqYvsDQ_cMe5h4AaABAg","responsibility":"ai_itself","reasoning":"deontological","policy":"none","emotion":"outrage"},
{"id":"ytc_Ugx7aWHk02Nh6tM30P94AaABAg","responsibility":"distributed","reasoning":"consequentialist","policy":"ban","emotion":"outrage"},
{"id":"ytc_UgzHbXSN-y1ouB8prpd4AaABAg","responsibility":"user","reasoning":"virtue","policy":"none","emotion":"approval"},
{"id":"ytc_UgzGhxeIqzYm3soFtjt4AaABAg","responsibility":"none","reasoning":"mixed","policy":"none","emotion":"mixed"},
{"id":"ytc_Ugxw0l9IxTYV5WsLcrt4AaABAg","responsibility":"developer","reasoning":"mixed","policy":"none","emotion":"outrage"}
]