“Are you a bird?” she asked.
The woman looked confused. “Hello, hello, I’m hungry, where am I?”
“A bird,” said Billie, and flapped her arms like wings.
The effect on Pitch was immediate. She sat up straighter on her perch and flapped her wings, not hard enough to take off, but hard enough to mimic the gesture.
“A bird!” announced the avatar. “A bird a bird a bird yes a bird. Are you a bird? Hello? A bird? Hello, can you hear me, hello?”
“Holy shit,” whispered Mike. “She’s really talking to the bird. The translation algorithm really figured out how to let her talk to the bird. And the bird is really talking back. Holy shit.”
“Not in front of my child, please,” I said, tone prim and strangled. The xenolinguists were going to be all over this. We’d have people clawing at the gates to try to get a place on the team once this came out. The science behind it was clean and easy to follow—we had built a deep neural net capable of learning, told it that gestures were language and that the human mouth was capable of making millions of distinct sounds, taught it to recognize grammar and incorporate both audio and visual signals into same, and then we had turned it loose, putting it out into the world, with no instructions but to learn.
“We need to put, like, a thousand animals in front of this thing and see how many of them can actually get it to work.” Mike grabbed my arm. “Do you know what this means? This changes everything.”
Conservationists would kill to get their subjects in front of a monitor and try to open communication channels. Gorillas would be easy—we already had ASL in common—and elephants, dolphins, parrots, none of them could be very far behind. We had opened the gates to a whole new world, and all because I wanted to talk to my sister.
But all that was in the future, stretching out ahead of us in a wide and tangled ribbon tied to the tail of tomorrow. Right here and right now was my daughter, laughing as she spoke to her new friend, the two of them feeling their way, one word at a time, into a common language, and hence into a greater understanding of the world.
Tasha would be so delighted.
In the moment, so was I.
New York Times best-selling author Seanan McGuire has enjoyed critical and commercial success both under her own name and the pseudonym Mira Grant. Heralded right out of the gate with 2010’s John W. Campbell Award for Best New Writer, McGuire also holds the unique honor of appearing five times on the same Hugo Awards ballot (with three nominations for McGuire and two for Grant). Prolific in both novels and short work, she is lauded for her detailed world-building in her October Daye urban fantasy series—currently planned out to at least thirteen entries that appear every September—and the InCryptid series, which kicked off with 2012’s Discount Armageddon. As Grant, she writes the bracing and topical Newsflesh trilogy. Its first volume, Feed, was one of Publishers Weekly’s Best Books of 2010, and earned nominations for both the Shirley Jackson Award and the Hugo Award for Best Novel.
Greg Bear
The Machine Starts
Though I am otherwise relentlessly normal, I have one peculiarity: I get along well only with people who are smarter than me. My wife, for example, is smarter than me. I am happy in my marriage.
In my present employment I should be very happy, because everyone around me is smarter and often at pains to prove that fact. It is my duty to reinforce their positive opinions, but at the same time to exert, now and then, small course corrections. Nothing shores up a fine self-opinion better than success.
So far, five years into our project, we had known nothing but failure.
The first thing you saw as you approached the perimeter site was the warehouse, large, square, and painted a brilliant titanium white. Surrounded by two high hurricane fences topped with glittering rolls of razor wire, it looked like the kind of place where you might store an A-bomb. Access to the site was on a strictly controlled, need-to-go basis. Parking was several hundred yards away, on a small lot covered with pulverized rubber. You were told not to drive a loud car, not to cut out your exhaust or rev your engine, not to sing or even shout, upon penalty of being fired.
On the morning of the test, I drove into the lot and parked my white VW, old and shabby. I had owned it since college. My colleagues favored Teslas or Mercedes-Benzes. I liked my Rabbit.
In the lane between the fences, small robots rolled night and day—nonlethal, but capable of shooting barb-tipped wires that carried a discouraging shock. The robots inspected me with their tiny black eyes and, bored by my familiarity, rolled away.
The warehouse was made entirely of wood, no nails or brackets. It covered half an acre and sat on a thick pad of cement reinforced with plastic rebar and mesh. Beneath the pad lay a series of empty vaults that discouraged ground water, rodents, or anything else that might disturb the peace. No pipes or wires were allowed, except for those that fed directly into the warehouse.
After I passed through the fences, a single thick oak door gave access to the warehouse interior. I was scheduled to meet Hugh Tiflin, project manager and chief researcher. He was always prompt, but I was deliberately early. I wanted to reacquaint myself with the architecture, the atmosphere, the implications—to feel the place again.
I summoned up my image of Alan Turing. It is my habit to sometimes talk to the founder of modern computing, hoping for a reflection of his peculiar, sharp wisdom. What we were in the final stages of creating (we all hoped they were the final stages!) could transform the human race. A machine that would end all our secrets. What would Mr. Turing think of such a New Machine?
He never answered, of course. But then, so far, neither did our machine.
I entered the security cage and listened to questions spoken by a soft, automated voice—personal questions that were sometimes embarrassing, sometimes sad, sometimes funny. I answered each of them truthfully enough and the cage opened.
Next to the cage, a small illuminated counter revealed the number of my recent visits: 4. In the last month, I had only been here twice. The counter reset every day. I take it as a personal affront when automated systems make mistakes.
A soft rain began to fall on the high, hollow roof, adding to my damp mood and the penetrating chill in the building. The warehouse was dark, except for a light in the far corner that glowed like a pale sun. I approached a low wood rail and stood in the long, curved shadow of a big black sphere, bloated and shiny, rising on tiny fins almost to the ceiling, silent but for the low hum of the power that kept it alive. A bank of heavily insulated pipes passed under the rails and through the wooden wall to dedicated generators and a refrigeration complex outside.
Early in its development, Tiflin had named the sphere Magic 8 Ball, soon shortened to 8 Ball because, as Tiflin insisted, there was nothing magical about our machine—just good solid physics. It retained a window on one side, however, like the old toy. Tiflin had asked it to be painted on after we finished the first phase.
The window’s message: TRY AGAIN LATER.
Reading that, I experienced an odd sort of dizzy spell, as if there were too many of me in one place—a symptom of stress and hard work, I presumed.
8 Ball was our third major attempt at a fully operational and manageable quantum computer. No doubt you’ve heard something about quantum computing. The underlying ideas are spooky and new, so a lot of what you’ve heard is bound to be wrong. A quantum computer works not with bits but with qubits, or quantum bits. A classic bit, like a light switch, is either on or off, one or zero. A quantum bit can be kept in superposition, neither on nor off, nor both, nor neither—like Schrödinger’s cat until you open its very special box.