When he kissed her, her lips tasted of salt. His last look at her was from fifteen hundred feet. She was a forlorn figure, standing out on the patio, waving listlessly.
He was a pale man, almost luminous in his pallor, and he announced himself as Roger Wald, Kayden’s Executive Assistant. Wald flapped his pale hands and Kayden thought that he looked as though his face was of moonstone dust, held together with luminous putty.
“How long have you been on the job here, Wald?”
“Oh, over two years. I’ve been the assistant to some very great men and—”
Kayden grinned. “Yeah. And now you’re the assistant to a guy with grease under his fingernails. Buck up, Roger. I brush my teeth and everything.”
Wald flapped his gray hands some more. “Oh, I didn’t mean to imply that—”
“Skip it, Roger. You just keep telling me the score and we’ll get along fine. Is this my room?” Wald had led him into a small plaster cubicle containing one single bed, a chair, a bureau and a glass ash tray.
“Yes, it is. I admit it’s a bit bleak, Mr. Kayden—”
“Call me Joe, please.”
“Yes sir. The room is bleak. They all are. Dr. Mundreath who was in charge three years ago felt that there should be no distractions, you know.”
“No, I don’t know. Let me check this. I’m in charge?”
“Oh, yes sir.”
“Then your first job, Roger, is to get me a suite of rooms. I want luxury on a Sybarite scale. I want rooms with music, tele sets, wine lists and everything but beautiful hostesses. Got it?”
“Yes, sir.”
“Now show me the production setup, the labs and all.”
The Project was housed in a series of long, one-story buildings surrounded by a high electrified wall. Interception rocket stations were set up in profusion in the surrounding countryside, the scanners revolving perpetually.
One building housed the best approach to a Thinking Machine that had been devised. The guard let them through the door and Kayden stopped dead. The main room was five hundred feet long and about eighty feet wide. All along the walls stood independent units of the machine. Each unit was plastered with switchboard panels, plug sockets and lamp indicators. Between the interstices of the panels showed an array of electronic tubes, circuit elements, relays.
Kayden looked at a small vehicle rolling smoothly across the floor. A uniformed girl sat in it and guided it. He recognized it as a massive variation of a master programming unit. The girl wheeled it up to one of the independent units against the wall, consulted a chart and plugged in the programming unit. The indicator lamps glowed and the girl took the tape that was ejected from the wall unit. She glanced at it, unplugged and wheeled away toward a far part of the room. He could see at least a dozen other master programming units. “What are they after?” he asked Wald.
“Test problem. With each improvement in the basic equipment, we run the same test problems through.”
“What’s the problem they’re working on now?”
Roger Wald beckoned to one of the girls on the vehicles. She stopped beside them, smiled prettily.
“Mr. Kayden, Miss Finch. Miss Finch, what is the test problem?”
“Chemical exchange separation method, Mr. Wald.”
The girl drove away on the silent wheels. Wald said: “We just feed the machine all the factors of a problem — i.e., to devise a simple way of preparing carbon-13 compounds. We know the answer, of course. Other test questions concern other fields — rules of harmonics, heat radiation and so forth.”
They walked into the room and, as Kayden looked more closely at the independent units, he began to see the point of approach to the problem. He said, “Give me a short statement of the reasons for failure.”
Roger Wald bit his lip. “My training... I’d better get Dr. Zander for you. He’s in charge of testing and analysis of results. We’ll go to his office.”
Zander was a man constructed of overlapping pink spheres. His face was covered with a constant dew of perspiration. He had the build, the complexion and the blue eyes to go with what should have been an amiable disposition. But his small mouth was an upside down U of sourness, his eyes were smothered bits of blue glass and his voice was a nasal whine. He looked at Kayden with what could have been contempt. Kayden sat and Wald stood on the opposite side of Zander’s paper-littered desk.
“So! You’re the new director,” Zander said.
“Right. Glad to know you, Dr. Zander. I’ve heard about you. Suppose you give me a brief on the present difficulties.”
“You want it in layman’s language?”
Kayden smiled with his lips alone. “I think I can struggle through the big words with you, Doc.”
Zander frowned and put his fat fingertips together, stared at Kayden through the puffy arch. “History first. By 1953 the Electronic Mechanical and Numerical Integrator and Calculator was carried to a point of development where it could solve any problem given to it in the mathematical field, provided the automatic sequencing was fed to it on a paper tape or punch cards. Iconoscopes were set up to act as accumulators to expand the memory factor, and calculations were put on the binary obviating the use of digits two through nine.
“With the first appropriation to develop a Thinking Machine, as it is called by the layman, our problem was to switch from mathematics to semantics. In other words, instead of absolute figures, we had to change over to the fuzzy values of words and phrases. Instead of asking for the cube root to ten thousand places of minus two, we had to ask it what happens when a cat is shot through the head and have it answer that the cat dies. As simple as that.
“To make the changeover, we had to select a language for it. We selected English and took out all variations which add little or nothing to connotation. We gave each sound a numerical value, and combined the numerical values into words. Then, into the expanded memory factor, we fed thousands of truisms. Naturally, with number-sound valuation, each truism became a formula... an equation. Assume that we had fed into the memory factor the phrase, ‘Roses are red.’ The machine tucks it away as a numerical formula. Then we ask the machine, ‘What color are roses?’ It translates the question into an open-ended formula, digs into the memory chamber and says back to us, ‘Roses are red.’
“Now we can ask a question based on any truism or proven statement that we have fed the machine, and we get the answer. We get it either written or spoken, though I personally consider the vocal attachments to be more toys than anything practical. The voice makes an impression on distinguished visitors, particularly when we permit the visitor to ask his own question. It is embarrassing when the question concerns a statement not previously fed to the memory factor. One congressman asked when his mother would die. The machine gave him a detailed definition of the word mother and a physiological explanation of the meaning of death — what happens when death occurs.
“The next step was to teach the machine basic differentiations. We selected a quality — such as calorie content. Then we stored in the memory factor a complete list of caloric ratings of food. Now, if you ask it the calorie rating of a given food, it will answer, or if you ask it which of two foods has the highest rating, it will select the proper answer. We have fed the machine eighty thousand differentiation lists covering eighty thousand different methods of grading myriad items.
“In addition,” he continued, “we have read to it philosophical concepts, records of phenomena, all types of data and information. At the present time we have a superabundance of response. Should you feed it just one word, such as ‘steel’ or ‘indigestion’ the machine will give you several volumes of data.”