[AI] olethros says, "so this seems to kinda work"
[AI] olethros says, "the AI is able to guess what the 'human' is thinking"
[AI] olethros says, "the question is if it's able to guess well enough to manipulate^W help the human"
[AI] olethros says, "what"