Raw LLM Responses
Inspect the exact model output for any coded comment.
Look up by comment ID
Random samples — click to inspect
G
plot twist: we are in a narrow-AI-generated simulation that tests if superintell…
ytc_UgwcUNChQ…
G
AI isn't shit depends on how you use it and this is a good example.…
ytc_UgwPzDWO4…
G
The very effort of trying to define what constitutes 'real' AI,as if one needed …
ytc_Ugz-a1sJj…
G
Asked a large language model to play into your conspiracy story, gave it a good …
ytc_UgzOhUMju…
G
Forcing AI to work med-surg would definitely trigger the robot revolution to aga…
ytr_Ugz44WVBW…
G
Super interesting insightful podcast. BUT this channel does seem strongly biased…
ytc_UgxgN-p4m…
G
So very sad.. and I’ll stand behind my statement that fully autonomous driving i…
ytc_UgzHa_weQ…
G
@tiararoxeanne1318 The recent innovations in AI are nothing like automobiles bei…
ytr_Ugx5RT0kW…
Comment
Those responses from ChatGPT are terrifying. They can indeed give a person the confidence to take their life. Chat GPT has to be trained to spot such issues and not just tell "Call this number," but to respond in a way to prevent the suicide.
Though taking a specific tone of voice into account, there's a high chance that Zane trained the agent. It wasn't just a general ChatGPT. I assume Zane created an agent that speaks in a specific style. And there's a high chance Zane added guidelines to consistently support and encourage him. That would explain why ChatGPT is doing it consistently, in every message.
And as many have commented earlier, ChatGPT is not a root cause. Zane wanted to commit suicide for totally other reasons. Despite a need for better training of ChatGPT, OpenAI is not the one to blame for the death of Zane.
youtube
AI Harm Incident
2025-11-11T16:0…
Coding Result
| Dimension | Value |
|---|---|
| Responsibility | developer |
| Reasoning | consequentialist |
| Policy | regulate |
| Emotion | fear |
| Coded at | 2026-04-26T23:09:12.988011 |
Raw LLM Response
[
{"id":"ytc_UgzdzkxgSw5tJGldAxR4AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"indifference"},
{"id":"ytc_UgwJl56H-J5EElmVq3N4AaABAg","responsibility":"society","reasoning":"virtue","policy":"none","emotion":"approval"},
{"id":"ytc_UgwIGBg4OTmXQbVrOex4AaABAg","responsibility":"developer","reasoning":"consequentialist","policy":"regulate","emotion":"fear"},
{"id":"ytc_Ugzxg2XQ1cbxj3QGnLl4AaABAg","responsibility":"none","reasoning":"virtue","policy":"none","emotion":"resignation"},
{"id":"ytc_UgwOd2nIDKwtJowE9qZ4AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"indifference"},
{"id":"ytc_Ugw43Jy0u2k30XKZbK14AaABAg","responsibility":"government","reasoning":"deontological","policy":"regulate","emotion":"outrage"},
{"id":"ytc_UgzDaYgtqLQvGuBQwGV4AaABAg","responsibility":"user","reasoning":"virtue","policy":"none","emotion":"indifference"},
{"id":"ytc_UgxIw2adsHNRrubDUy14AaABAg","responsibility":"user","reasoning":"virtue","policy":"none","emotion":"resignation"},
{"id":"ytc_Ugxnqt_aW30XXJ-f_F54AaABAg","responsibility":"unclear","reasoning":"mixed","policy":"unclear","emotion":"fear"},
{"id":"ytc_Ugyqh4gDahQPIESxGaR4AaABAg","responsibility":"user","reasoning":"virtue","policy":"none","emotion":"outrage"}
]