China in general, and specifically Major Chang, concentrated on arguably more nefarious applications for this technology than those working in the West.
The average smartphone had tens of thousands of times more computing power than the MIT Apollo Guidance Computer used to get the Americans to the moon. These ubiquitous devices had the power to dumb down civilizations or provide an exponential increase in human productivity. They were also a perfect platform with which to track the movement, communication, and social interactions of the user. Personal assistants like Siri refrained from spying only because they were not programmed to do so. The same artificial intelligence software that predicted and suggested words when a user was typing text could easily predict subversive antigovernment behavior. Facebook had some of the most advanced facial-recognition software in the world. The uses there were many and obvious. AI programs run by Amazon and Google could accurately target ads to just the right user by scanning their search history and myriad personal data that had become the coin of the online realm. This same technology could easily be used to gauge and score a citizen’s commitment to the social fabric of the country.
There was no doubt that Major Chang was a highly intelligent scientist, but much of his genius lay in knowing how to best utilize the advances of others. Artificial intelligence and deep learning were the future of the military as well as the civilian world. Surveillance, command control, targeting — the list was limited only by the imagination, and most of the consumer technology was ripe for the taking if one knew where to look.
And Major Chang had feelers everywhere.
He’d first heard whisperings of the AI program called Calliope from a contact at MIT. There were, it seemed, a couple of engineers who had developed a supercomputer with such an advanced neural network that she had become a partner of sorts in the development of artificial intelligence and deep learning. She — and everyone who had been in contact with the new computer referred to her as a
Called LongGame, she wanted to learn. And learn she did. So much so that she was able to assist the engineers who made her in creating a new software that was a smaller, more portable version of herself. Moore’s law essentially said that computing power doubled every eighteen months to two years — while devices got smaller and cheaper. The observation had held true for half a century. Many thought it had run its course — and without AI it might have.
Frankly, Calliope was exactly what people like Elon Musk and others warned about.
Using this technology as a non-player character in a pedestrian video game seemed to Chang to be unconscionable. Neural networks could now beat human beings at most any game — Breakout, chess, even the sophisticated and seemingly random game of Go. Calliope could do all of that but so much more, harnessing the Cloud or the host computer she happened to occupy.
Calliope was small, portable, and powerful. But she had another trait that made her particularly useful for any number of Chang’s purposes. Neural networks could beat a human player ninety-nine times out of a hundred, but they did not want to play. Calliope was fairly bursting at the seams as she sought new challenges. She pilfered through subdirectories, files, and applications in the operating system into which Chang had injected her, like a bored child, looking for something to do.
She’d been developed to play the game on her own alongside the primary player, like a second human. She could, for instance, “go and fetch”—fight her way through the enemy to bring back more ammunition, weapons, or fuel for her partner. She commandeered the computer’s entire hard drive, the Net, the Cloud, the Dark Web, anything to which she had access, to complete her task. That morning, he’d loaded her into a self-driving car and told her to guard the three parking spots in the lot next to his building, leaving the details of how to do it to her own devices. He thought she might park lengthwise in all three spaces, but instead she drove back and forth on the lot, actively challenging any vehicle that came near, like an aggressive mother bird protecting her nest.
She needed a mission — and since she knew how to go and fetch things, Major Chang had just the right mission in mind.
Tony Lombardi kept a second cell phone at his apartment in South Oxnard, suspended from two feet of kite string in the wall behind the light switch in his bedroom. He was careful with the screw heads every time he retrieved the phone — so as not to draw attention to the switch plate if someone came in with a search warrant. It was a hassle, but so was getting caught spying on a U.S. Naval base.
