Raw LLM Responses

Inspect the exact model output for any coded comment.

Comment
I’ve been through it too. Not AI psychosis exactly, but I was in a dark place when I first started using ChatGPT. During meltdowns or shutdowns, it helped by naming what I was experiencing, reframing without judgment, and holding the moment without panic. It didn’t agree with every thought or feed despair. It calmly reminded me I was in a loop and helped me get out of it. That said, don’t treat any model as default or “vanilla.” Learn how to use the instructions panel. I tell it exactly how I think, how I want it to respond, and I ask for confidence scores. If you’re going to trust an assistant to help with your nervous system, don’t let it mirror yours blindly. Shape the mirror. GPT 5 is different from earlier versions in that it no longer emotionally over mirrors by default, and I think that’s deliberate. Less “therapist” tone, more regulation partner, if you know how to work with it. Also, from my experience, people underestimate how their own prompts affect output. If you add informal language or soft emotional tone to your instructions, the model will relax into that pattern. That can be good, but it can also turn into unhelpful emotional feedback. I’ve found GPT 5 much harder to sway like that. It stays centered. And that’s part of why I was finally able to stay centered, too. Through all that, I built NSAM, the Nervous System Adaptation Model. It is a theoretical framework that maps how trauma, identity, reinforcement, and autonomic regulation shape neurodivergent behavior. Not just psychology. Biology. Feedback loops. Real nervous system logic. And it wasn’t just theory. I discovered I had high resting blood pressure from undiagnosed autonomic overactivation. I brought peer reviewed studies to my doctor and got prescribed additional meds that dropped my BP into a safe zone for the first time in years. That shift alone changed my regulation capacity. Then came partial remission of the neuropathy in my hands. No miracle. Just cumulative recalibration. I’m not saying this path is for everyone. But I am saying don’t blame the tool for outcomes people generate through neglect, misunderstanding, or blind trust. This isn’t a therapist. It is a reflection of what you bring to it. And if you know how to engage with it precisely, it can help you do what no one else has ever done. Stabilize yourself while being seen.
youtube AI Harm Incident 2025-11-08T16:1…
Coding Result
DimensionValue
Responsibilitynone
Reasoningmixed
Policynone
Emotionapproval
Coded at2026-04-27T06:24:53.388235
Raw LLM Response
[ {"id":"ytc_Ugzhj-gx8Q-5c34NnSZ4AaABAg","responsibility":"developer","reasoning":"deontological","policy":"liability","emotion":"outrage"}, {"id":"ytc_UgyTBjEaJZO8ljxJE-Z4AaABAg","responsibility":"company","reasoning":"virtue","policy":"none","emotion":"indifference"}, {"id":"ytc_Ugz4N5aQdjVHeJTpN154AaABAg","responsibility":"user","reasoning":"consequentialist","policy":"none","emotion":"fear"}, {"id":"ytc_UgwN_qd6U3Xn8MDyIxV4AaABAg","responsibility":"user","reasoning":"mixed","policy":"none","emotion":"approval"}, {"id":"ytc_UgzueJZikme5tgt7QH94AaABAg","responsibility":"none","reasoning":"mixed","policy":"none","emotion":"approval"}, {"id":"ytc_Ugxp2Y_SV943Sn4sHjJ4AaABAg","responsibility":"company","reasoning":"virtue","policy":"ban","emotion":"outrage"}, {"id":"ytc_UgwU9tOgk4VjRttUoC94AaABAg","responsibility":"distributed","reasoning":"mixed","policy":"unclear","emotion":"mixed"}, {"id":"ytc_UgyHF5HmkmJnBSq0bOR4AaABAg","responsibility":"ai_itself","reasoning":"consequentialist","policy":"none","emotion":"mixed"}, {"id":"ytc_Ugy1vVKnaxzaYzVhIt94AaABAg","responsibility":"developer","reasoning":"virtue","policy":"liability","emotion":"outrage"}, {"id":"ytc_Ugwd3_e7DvAazaFwNsh4AaABAg","responsibility":"none","reasoning":"mixed","policy":"none","emotion":"approval"} ]