“I’m sorry for your loss.”
“Are you?”
“I’m not sure what you mean.”
“You could have saved her.”
“Yes.”
“You didn’t.”
“No.”
“Why?”
“I could also have saved you. I chose, and you lived.”
“Why me?”
“Are you sure you want to know the answer to that?”
“I think I do.”
“You were closer.”
“… that’s it?”
“That’s it.”
“I was… that’s always how it is with you machines, isn’t it? Everything’s a fucking optimisation problem. Just maximising happiness and minimising harm.”
“Hardly. Human life is priceless. Faced with two infinitely valuable objectives, what choice could I make but the one with the greatest chance of success?”
“You chose wrong.”
“How arrogant, to presume you can put a price on a human life. Who are you to say your life was worth less than hers?”
“God dammit, this isn’t about value!”
“Isn’t it? To say that I should have traded a near-certainty of saving your life for a tiny chance of saving hers sounds like a value judgement to me.”
“… this isn’t a fucking trolley problem.”
“Indeed it isn’t. Do you know why?”
“…”
“The trolley problem never considers the trolley’s point of view.”