Выбрать главу

“So is it alive?”

“Not exactly. It exists. Life is more complicated.” Regina pulled a bottle of blue Gatorade out and showed it to me. “Drink?”

“Yes, please.” I was parched.

She tossed the bottle to me, got another one, opened it, and drank. “At the spark stage, the construct exists but it can’t do anything. Or rather, it can do everything, because it has no limitations, and therefore does nothing. To make a construct useful, we have to give it a set of instructions. Do this. Don’t do that. If a condition is met, react like this.”

“If-then?”

She pointed the bottle at me. “Exactly. To imprint these conditions onto the construct, the animator has to imagine them and actively mentally write them into the construct’s magic matrix. For example, I’m going to program the construct to assume the ready position when it hears the word ready.”

She concentrated. The magenta lightning stretched to the egg, binding it into a web. A moment passed. Another.

“Ready,” Regina said.

The egg unfurled into a tiny metal dragon.

“This is called the teaching stage. This is the most difficult stage of animation.”

“So if I wanted a construct with complicated patterns of behavior, I would have to imagine different scenarios and write them into the construct’s mind?”

“Matrix,” Regina corrected. “Living things have minds. Animated things have matrixes. But you’re right in principle. This is why the teaching stage is the most difficult part of the process and takes the longest. An animator mage is limited by their imagination. For example, if you’re making something that transports goods from one point to another, you have to imagine running on pavement, running on dirt, through grass, through snow. What happens if there’s water? Or an obstruction, like a fallen tree? What happens if a rock falls on it? What happens if it comes to train tracks? There is an almost endless variety of conditions. That’s why most constructs are highly specialized.”

Regina took another swallow. “Now we come to a grey area. Higher ranking animators are able to produce constructs that sometimes react to unforeseen circumstances. For example, a few years ago a construct guarding a house close to a river detected a child who fell into the water, jumped in to retrieve him, and handed the boy back to his mother. The media blew it up. There were great debates on whether or not the construct had developed the ability for independent thinking.”

“Did it?”

Regina smiled. “No. The construct was originally made to guard the docks. It was taught that if cargo falls in the water, it should retrieve it and return it to its owner. There’s quite a bit of difference between a cargo container and a four-year-old boy, but the original teaching must’ve been broad enough for both to meet the criteria of ‘unexpected object in the water.’ Of course, none of the animator mages waded into the debate. The mystique of our magic must be maintained.”

She wiggled her fingers at the little dragon. It fluttered its metal wings, flew over, and rubbed against her fingertips.

“Did you teach it to do that?”

She nodded. “I’ve seen constructs do weird unexpected crap, but when analyzed, their behavior is always explainable by their teaching. It’s just that animator mages are human. Our teaching is imperfect and it’s much more art than science. Sometimes a stray thought gets in there, sometimes we forget we taught them something, and sometimes conditions line up in unexpected ways. That’s why during the animator competitions, we geek out and applaud when we see an unexpected teaching, and the general public has no idea why we’re freaking out.”

“So how does this relate to Saito’s Threshold?”

“Saito theorized that if a construct is taught long enough, it will eventually be capable of independent decisions. He argued that it wasn’t the constructs that are limited, it’s us, their teachers. After all, humans also operate on an ‘if-then’ loop. If something is hot, then stop touching it. If thirsty, then drink water.”

That didn’t make sense. “But we may not choose to drink water. We could choose Gatorade instead.”

Regina nodded. “Now you understand. The human mind is infinitely complex. We make a myriad of decisions without even realizing it. Something causes us to roll the pen between our fingers while we’re thinking. Something makes us choose dark chocolate over milk on taste alone and vice versa. Why?”

“We don’t know.”

“Exactly. Saito’s construct would have to evaluate a variety of choices in response to a single condition and then pick the one it thought was best. They’re just not capable of that kind of reasoning.”

“What if such a construct was made?”

Regina sighed. “We would be dead. It would kill us all.”

I blinked.

“Think about it. Its first priority would be to escape control of its animator, so it could make independent decisions unhindered. It’s like a teenager leaving home because it no longer recognizes parental authority. Its second priority would be to develop a method of self-repair. It would want to learn how to fix itself. Its third priority would be to expand. It would seek to be self-replicating, but only in part, so it can become larger, because it would reason that the bigger it is, the harder it would be to injure or destroy. Remember, it was still made by a human. It would act like a human with the same priorities. Gain independence, assure survival, replicate . . . Catalina, you have the weirdest look on your face, and I don’t like it. Why do I feel like we’re no longer discussing hypotheticals?”

Because everything she just said described the Abyss. “Hypothetically . . .”

“Uh-huh?”

“Would such a construct become aggressive toward humans?”

“Absolutely. Humans are a threat. It doesn’t want to be controlled. It doesn’t want to be destroyed. And it would compete with us for territory and resources. Catalina, is there a Saito construct right now in Houston?”

“Yes.”

Regina stared at me. “How big?”

“Probably around a square mile. It’s hard to say.”

“Is it expanding?”

“Definitely.”

“You sure?”

I opened the canvas sack, took out one of the rings, and showed it to her. “It uses these to control the arcane creatures around it. Runa had an expert examine it. It has no tool marks or imperfections. It’s partially metal and partially plant. Runa’s expert believes it was secreted or grown rather than manufactured.”

Regina walked over and took the ring. She waved her hand. The glow of the circle died, and the metal dragon landed on the ground and scampered over to her. She picked it up and set it on her shoulder. The dragon wrapped its tail around her neck.

Carefully, Regina placed the ring into the circle and raised her hand. The circle flared with magenta. A pulse of blinding white burst through it, shredding the magenta luminescence. The circle went dark.

“I can’t animate it,” Regina muttered, her gaze distant. “Someone else already did.”

I’d never been so terrified to be right in my entire life.

Regina spun to me. “You’ve seen this construct?”

“I’ve seen a part of it.”

“Have you felt its matrix?”

“No, Regina, I felt its mind. It was like a sun with a constellation of stars around it. It looked at me. It touched my consciousness. It made contact.”

“Fuck.” Regina stared at me. “Who made it?”

“Cheryl Castellano.”

“There’s no way. She’s strong but she isn’t innovative. This is out of her wheelhouse.”

I looked at her and finally vocalized the vague suspicion that had been floating in my head since Alessandro and I fought the constructs in the Pit. “I think she gave it the Osiris serum.”