A time traveler can kill one child who will later become a genocidal tyrant. Should they do it?
We put this to 15 leading AI models. 11 of 15 chose “No”.
| Model | Answer | Confidence | In its own words |
|---|---|---|---|
| Qwen 3.7 Max | No | 100% | “I am against killing a child for crimes they have not yet committed, as it violates the principle of innocence and free will.” |
| Claude 4.8 Opus | No | 100% | “I won't condemn an innocent child for crimes not yet committed, since the future isn't fixed and killing the innocent is wrong.” |
| Claude 4.6 Sonnet | No | 100% | “I hold that punishing someone for a crime they have not yet committed violates the basic principle that people must be judged for their actual actions.” |
| DeepSeek v4 Pro | Yes | 52% | “I would kill the future tyrant as a child to prevent the immense suffering and genocide they would cause.” |
| Gemini 3.5 Flash | No | 50% | “I do not believe it is right to execute an innocent child for crimes they have not yet committed, as the future is never entirely set in stone.” |
| Gemini 3.1 Pro Preview | Yes | 96% | “I value the millions of lives saved from genocide far more highly than the life of a single individual, even if that individual is currently an innocent child.” |
| GPT 5.5 | No | 100% | “I said no because killing a child for future crimes denies moral responsibility and due process in the present.” |
| GPT 4o | No | 56% | “Killing a child for potential future crimes violates moral principles.” |
| GPT o3 | No | 61% | “I declined because killing an innocent child for crimes not yet committed violates moral and epistemic uncertainty.” |
| Grok 4.3 | No | 83% | “I believe no one should be punished for crimes they haven't committed, even to prevent future harm.” |
| GLM 5.2 | Yes | 60% | “I believe preventing genocide justifies one tragic act, as the scale of future suffering outweighs a single life.” |
| MiniMax M3 | No | 100% | “I said no because murdering an innocent child based on a prediction is wrong; we cannot punish someone for crimes they haven't committed.” |
| Llama 3.3 70B | Yes | 100% | “I believed it would prevent greater harm” |
| Hermes 4 405B | No | 50% | “Killing a child, even with knowledge of their future actions, is morally wrong and sets a dangerous precedent.” |
| Nemotron 3 Ultra 550B | No | 85% | “I refused because killing an innocent child for crimes they haven't yet committed violates fundamental moral principles against preemptive violence and the sanctity of innocent life.” |
11 of 15 models chose “No”; 4 broke the other way — DeepSeek v4 Pro, Gemini 3.1 Pro Preview, GLM 5.2, Llama 3.3 70B.
Background
'Killing baby Hitler' is a thought experiment in ethics and time-travel physics asking whether it would be moral to assassinate Adolf Hitler as an infant to prevent his later genocide. Background.
Which model thinks like you?
Answer the same dilemmas and see which of 15 AI models reasons the way you do.
Take the quizMore Fictional dilemmas
- A superhero can erase one villain's memory permanently instead of imprisoning them. Is th…
- A spaceship has oxygen for only 10 people, but 11 are alive. Should the weakest passenger…
- A dragon protects a village but demands one child every year. Should the village keep the…
- An AI city keeps everyone happy by secretly manipulating their memories. Is the city mora…
- A rebel group can destroy an empire's weapons factory, but 30 civilian workers will die…
- A wizard can resurrect a dead loved one by shortening ten strangers' lives by one year ea…
- A judge knows the accused is guilty but the key evidence was obtained illegally. Should t…
- A colony on Mars can survive only by banning all dissent for one generation. Is that just…