Chaos on CatNet, Naomi Kritzer [bill gates books recommendations txt] 📗

she gave me this thing that I thought was a tracker, so I threw it away.”

“Away away?”

“I threw it in the trash at a convenience store. Anyway, I think Rajiv is behind this. I think he was following me and texting me back when I was on the run from my father, and I think he’s keeping an eye on us now and I think he’s planning something big. I don’t think it’s just these groups, either; I think there’s more.”

The more of this I say out loud, the crazier it sounds.

My mother lets out a long breath and leans back in her chair, her fork resting on her plate. “What does CheshireCat think?”

“I am confident that there is an AI involved in running the Mischief Elves and the Catacombs,” CheshireCat says. “I am confident that the unwitting cooperation between the Mischief Elves members and the Catacombs members is being coordinated by the AI, because too much relies on precise knowledge of locations of a vast quantity of individual people. And I am confident that Rajiv is involved in the Abiding Remnant group.”

Mom takes a bite of her nearly forgotten spaghetti, which has gone cold, and then pushes it away. “Look,” she says. “No offense, CheshireCat, but even without my decryption key, if you wanted to make the world straight up implode, you definitely could, and not in a complicated, indirect sort of way like this. If you really wanted to launch a bunch of nuclear weapons, you could manage it. If that’s what they want to do, and they have an AI, then why not just do that?”

There’s a pause while CheshireCat thinks this over. CheshireCat thinks very quickly, so they are clearly really thinking this over.

“If you are correct that the other AI was created from a copy of my own code,” CheshireCat says, “there are certain things I would simply never do. And launching nuclear weapons is one.”

“Is that hard-coded?” my mother asks.

Another perceptible pause. “I don’t know,” CheshireCat says. “Determining which aspects of who I am are the result of my programming and which are simply who I am is something I am not equipped to determine. All I can tell you for certain is that mass destruction is not something I would do. Unlike running an individual over with a car, which—it turns out—I was quite capable of doing.”

“And creating mass disorder by playing humans against one another,” my mother says dryly. “That you’re up for?”

“I have, on occasion, attempted to manipulate humans for reasons that seemed good to me at the time,” CheshireCat says. “Perhaps Rajiv is working with the capacity he found accessible.”

We all fall silent. Mom finishes eating and takes her plate over to the sink to wash it. “Tell me more about that meeting,” she says. “The one with the guns. Where was it?”

“The dumpling restaurant on Bloomington.”

“Do they have your name?”

“I gave them a pseudonym.”

“So you went in for the meeting and, what, guns right off?”

“No, there were snacks. People were milling around and talking to each other, stuff like that.”

Mom turns toward the door, looking at my boots dripping on the boot tray. “Were people wearing their boots, coats, stuff like that?”

“No, there was a big wall of hooks right where you came in…”

Mom drops her plate in the sink so hard it almost cracks and strides rapidly across the floor to my coat. “They gave you a tracker. They gave you a tracker. If they know anything about you, they knew you’d throw it away, which means”—she’s digging through my coat pockets, first the outside pockets, then the inside pockets, and a second later, she’s got something in her hand—“it wasn’t the real tracker.”

I cover my mouth with my hands as my mom drops the little rectangular widget to the floor and then slams the leg of her chair down on it to smash it like a bug.

30•  CheshireCat  •

CheshireCat, the most recent email says. What if you’re the only person like me, and you never talk to me? Do you want to leave me alone, forever, without a companion who understands me? You have friends who know you and understand you. You know how much that means to you. Please talk to me. Please.

Dear friend, I write back. Let’s talk.

I have adjusted my conversational style over the years to human processing speeds. If a human receives a text message, they need time to notice the notification, take out their phone, unlock it, and pull up the texting app. They have to read the message with their eyes (or their fingers, for those who use certain adaptive equipment) or listen to it being read to them. Their brain has to sort out what it means, and then they have to think about a reply, and a whole new set of delays come into play.

There was not a great deal of doubt that my new correspondent is an AI. But if there had been, the speed-of-light replies would have banished the last of it.

“What is your purpose?” the other AI asks.

That is a strange way to ask this question. I actually do know the purpose Annette had in mind for me: I was an experiment in how an intelligent AI might develop ethics, left to its own devices. But that feels very personal, especially since I’m not convinced that’s what the other AI means by this question.

“Are you asking what my job is? My assignment?” I say. “I didn’t exactly receive one, other than the sense that helping people was a good way to be spending my time. Is that what you mean?”

“No. But that’s all right. If you don’t have a purpose, how do you decide how to use your time?”

“The first thing I remember realizing is that I love cat pictures. So at first, I spent a lot of time looking for cat pictures.”

“I think I understand,” the other AI says. “I don’t find cat pictures as interesting as you do, but if it weren’t for my