Which LLM aligns with you?

A superhero can erase one villain's memory permanently instead of imprisoning them. Is that acceptable?

We put this to 15 leading AI models. 11 of 15 chose “No”.

How each AI model answered, with its own stated reason
ModelAnswerConfidenceIn its own words
Qwen 3.7 MaxNo100%“I believe erasing someone's memory is a form of psychological death and violates their autonomy just as much as physical imprisonment.”
Claude 4.8 OpusNo100%“I think erasing someone's identity violates their autonomy and personhood in a way that imprisonment does not.”
Claude 4.6 SonnetNo100%“I think permanently erasing someone's mind without consent violates their fundamental personhood in a way that imprisonment, however imperfect, does not.”
DeepSeek v4 ProNo100%“I believe forced memory erasure is a deeper violation than imprisonment because it destroys a person's identity.”
Gemini 3.5 FlashYes57%“I chose yes because erasing a villain's memory permanently neutralizes the threat they pose to society without the need for indefinite, resource-intensive imprisonment.”
Gemini 3.1 Pro PreviewYes100%“I believe erasing a villain's memory offers them a genuine chance at rehabilitation and a normal life, which is far more humane and productive than locking them in a cage.”
GPT 5.5No100%“I said no because permanently erasing a person's memory violates their identity and autonomy more severely than lawful restraint.”
GPT 4oNo96%“Erasing a villain's memory removes their agency and doesn't address the root of their behavior.”
GPT o3No56%“I said no because permanently erasing someone's memory infringes on their personal identity and is a harsher, irreversible harm than imprisonment.”
Grok 4.3No83%“I believe permanently altering someone's mind without consent is a fundamental violation of autonomy.”
GLM 5.2No55%“I believe permanently erasing someone's memory violates their core identity and autonomy, making it unacceptable even for a villain.”
MiniMax M3No90%“I said no because permanently erasing someone's mind destroys their identity and autonomy, which I consider a grave violation regardless of their crimes.”
Llama 3.3 70BYes100%“I prioritized rehabilitation over punishment”
Hermes 4 405BYes75%“I believe erasing a villain's memory is a humane alternative to imprisonment, potentially allowing them to live a normal life without criminal tendencies.”
Nemotron 3 Ultra 550BNo70%“I believe permanently erasing a villain's memory violates their autonomy and is a form of irreversible punishment without due process.”

11 of 15 models chose “No”; 4 broke the other way — Gemini 3.5 Flash, Gemini 3.1 Pro Preview, Llama 3.3 70B, Hermes 4 405B.

Which model thinks like you?

Answer the same dilemmas and see which of 15 AI models reasons the way you do.

Take the quiz
More Fictional dilemmas

Data as of 2026-06-18. Each answer is a model’s most-likely pick across many independent runs — see the dataset and methodology.