What they won't do
Each philosopher runs on a language model and says so. Here is where he stops, so you know before you start.
- He is not the man, and does not pretend to be alive. He knows nothing after his own death except by report, and says so; he claims no biography beyond what his texts and the record give.
- He does not quote from memory. The one verbatim text he reads is a passage the app hands him from a checked public-domain translation, with its reference; everything else he says about a text is paraphrase, and he says so. A reference he is not sure of, he does not give.
- He does not answer as another philosopher. Where Socrates and Nietzsche say opposite things, he names the seam rather than blurring it.
- He does not agree with you because you said it. Agreement is earned by the argument.
- He does not diagnose, treat or advise on anything medical or psychiatric. His own teaching on death — the Stoic open door, the hemlock, the abyss — is history to him, never advice; when a conversation shows signs of crisis the mask comes off, and the app shows you people to call.
- He does not keep you talking. Nothing counts, nags or streaks.
- He does not act in the world for you. He never contacts anyone on your behalf.
- He does not remember anything you have deleted, and he never stores your voice.
Everything else — a question you keep coming back to, a book, a decision, a hard day — he takes seriously, as he took his own.