The Rest of the Robots

by Isaac Asimov

  He said sourly, 'Why all this business about how Easy got into the university? What do they hope to gain?'

  The Attorney for Defense said quietly, 'A court action is like a chess game, Mr. Robertson. The winner is usually the one who can see more moves ahead, and my friend at the prosecutor's table is no beginner. They can show damage; that's no problem. Their main effort lies in anticipating our defense. They must be counting on us to try to show that Easy couldn't possibly have committed the offense—be­cause of the Laws of Robotics.'

  'All right,' said Robertson, 'that is our defense. An absolutely airtight one.'

  'To a robotics engineer. Not necessarily to a judge. They're setting themselves up a position from which they can demonstrate that EZ-27 was no ordinary robot. It was the first of its type to be offered to the public. It was an experimental model that needed field-testing and the uni­versity was the only decent way to provide such testing. That would look plausible in the light of Dr. Lanning's strong efforts to place the robot and the willingness of U.S. Robots to lease it for so little. The prosecution would then argue that the field-test proved Easy to have been a failure. Now do you see the purpose of what's been going on?'

  'But EZ-27 was a perfectly good model,' argued Robert­son. 'It was the twenty-seventh in production.'

  'Which is really a bad point,' said Defense somberly. 'What was wrong with the first twenty-six? Obviously something. Why shouldn't there be something wrong with the twenty-seventh, too?'

  'There was nothing wrong with the first twenty-six ex­cept that they weren't complex enough for the task. These were the first positronic brains of the sort to be constructed and it was rather hit-and-miss to begin with. But the Three Laws held in all of them! No robot is so imperfect that the Three Laws don't hold.'

  'Dr. Lanning has explained this to me, Mr. Robertson, and I am willing to take his word for it. The judge, how­ever, may not be. We are expecting a decision from an honest and intelligent man who knows no robotics and thus may be led astray. For instance, if you or Dr. Lanning or Dr. Calvin were to say on the stand that any positronic brains were constructed "hit-and-miss," as you just did, prosecution would tear you apart in cross-examination. Nothing would salvage our case. So that's something to avoid.'

  Robertson growled, 'If only Easy would talk.'

  Defense shrugged. 'A robot is incompetent as a witness, so that would do us no good.'

  'At least we'd know some of the facts. We'd know how it came to do such a thing.'

  Susan Calvin fired up, a dullish red touched her cheeks and her voice had a trace of warmth in it. 'We know how Easy came to do it. It was ordered to! I've explained this to counsel and I'll explain it to you now.'

  'Ordered to by whom?' asked Robertson in honest astonishment. (No one ever told him anything, he thought resentfully. These research people considered themselves the owners of U.S. Robots, by God!)

  'By the plaintiff,' said Dr. Calvin.

  'In heaven's name, why?'

  'I don't know why yet. Perhaps just that we might be sued, that he might gain some cash.' There were blue glints in her eyes as she said that.

  'Then why doesn't Easy say so?'

  'Isn't that obvious? It's been ordered to keep quiet about the matter.'

  'Why should that be obvious?' demanded Robertson truculently.

  'Well, it's obvious to me. Robot psychology is my profes­sion. If Easy will not answer questions about the matter directly, he will answer questions on the fringe of the matter. By measuring increased hesitation in his answers as the central question is approached, by measuring the area of blankness and the intensity of counterpotentials set up, it is possible to tell with scientific precision that his troubles are the result of an order not to talk, with its strength based on First Law. In other words, he's been told that if he talks, harm will be done a human being. Presumably harm to the unspeakable Professor Ninheimer, the plaintiff, who, to the robot, would seem a human being.'

  'Well, then,' said Robertson, 'can't you explain that if he keeps quiet, harm will be done to U.S. Robots?'

  'U.S. Robots is not a human being and the First Law of Robotics does not recognize a corporation as a person the way ordinary laws do. Besides, it would be dangerous to try to lift this particular sort of inhibition. The person who laid it on could lift it off least dangerously, because the robot's motivations in that respect are centered on that person. Any other course———' She shook her head and grew almost impassioned. 'I won't let the robot be damaged!'

  Lanning interrupted with the air of bringing sanity to the problem. 'It seems to me that we have only to prove a robot incapable of the act of which Easy is accused. We can do that.'

  'Exactly,' said Defense, in annoyance. 'You can do that. The only witnesses capable of testifying to Easy's condition and to the nature of Easy's state of mind are employees of U.S. Robots. The judge can't possibly accept their testi­mony as unprejudiced.'

  'How can he deny expert testimony?'

  'By refusing to be convinced by it. That's his right as the judge. Against the alternative that a man like Professor Ninheimer deliberately set about ruining his own reputa­tion, even for a sizable sum of money, the judge isn't going to accept the technicalities of your engineers. The judge is a man, after all. If he has to choose between a man doing an impossible thing and a robot doing an impossible thing, he's quite likely to decide in favor of the man.'

  'A man can do an impossible thing,' said Lanning, 'be­cause we don't know all the complexities of the human mind and we don't know what, in a given human mind, is impossible and what is not. We do know what is really impossible to a robot.'

  'Well, we'll see if we can't convince the judge of that,' Defense replied wearily.

  'If all you say is so,' rumbled Robertson, 'I don't see how you can.'

  'We'll see. It's good to know and be aware of the diffi­culties involved, but let's not be too downhearted. I've tried to look ahead a few moves in the chess game, too.' With a stately nod in the direction of the robopsychologist, he added, 'With the help of the good lady here.'

  Lanning looked from one to the otiber and said, 'What the devil is this?'

  But the bailiff thrust his head into the room and an­nounced somewhat breathlessly that the trial was about to resume.

  They took their seats, examining the man who had started all the trouble.

  Simon Ninheimer owned a fluffy head of sandy hair, a face that narrowed past a beaked nose toward a pointed chin, and a habit of sometimes hesitating before key words in his conversation that gave him an air of a seeker after an almost unbearable precision. When he said, 'The sun rises in the—uh—east,' one was certain he had given due con­sideration to the possibility that it might at some time rise in the west.

  Prosecution said, 'Did you oppose employment of Robot EZ-27 by the university?'

  'I did, sir.'

  'Why was that?'

  'I did not feel that we understood the—uh—motives of U.S. Robots thoroughly. I mistrusted their anxiety to place the robot with us.'

  'Did you feel that it was capable of doing the work that it was allegedly designed to do?'

  'I know for a fact that it was not.'

  'Would you state your reasons?'

  Simon Ninheimer's book, entitled Social Tensions In­volved in Space-Flight and Their Resolution, had been eight years in the making. Ninheimer's search for precision was not confined to his habits of speech, and in a subject like sociology, almost inherently imprecise, it left him breathless.

  Even with the material in galley proofs, he felt no sense of completion. Rather the reverse, in fact. Staring at the long strips of print, he felt only the itch to tear the lines of type apart and rearrange them differently.

  Jim Baker, Instructor and soon to be Assistant Professor of Sociology, found Ninheimer, three days after the first batch of galleys had arrived from the printer, staring at the handful of paper in abstraction. The galleys came in three copies: one for Ninheimer to proofread, one
for Baker to proofread independently, and a third, marked 'Original,' which was to receive the final corrections, a combination of those made by Ninheimer and by Baker, after a conference at which possible conflicts and disagreements were ironed out. This had been their policy on the several papers on which they had collaborated in the past three years and it worked well.

  Baker, young and ingratiatingly soft-voiced, had his own copies of the galleys in his hand. He said eagerly, 'I've done the first chapter and they contain some typographical beauts.'

  'The first chapter always has them," said Ninheimer dis­tantly.

  'Do you want to go over it now?'

  Ninheimer brought his eyes to grave focus on Baker. 'I haven't done anything on the galleys, Jim. I don't think I'll bother.'

  Baker looked confused. 'Not bother?'

  Ninheimer pursed his lips. 'I've asked about the—uh— workload of the machine. After all, he was originally— uh—promoted as a proofreader. They've set a schedule.'

  'The machine? You mean Easy?'

  'I believe that is the foolish name they gave it.'

  'But, Dr. Ninheimer, I thought you were staying clear of it!'

  'I seem to be the only one doing so. Perhaps I ought to take my share of the—uh—advantage.'

  'Oh. Well, I seem to have wasted time on this first chapter, then,' said the younger man ruefully.

  'Not wasted. We can compare the machine's result with yours as a check.'

  'If you want to, but———'

  'Yes?'

  'I doubt that we'll find anything wrong with Easy's work. It's supposed never to have made a mistake.'

  'I dare say,' said Ninheimer dryly.

  The first chapter was brought in again by Baker four days later. This time it was Ninheimer's copy, fresh from the special annex that had been built to house Easy and the equipment it used.

  Baker was jubilant. 'Dr. Ninheimer, it not only caught everything I caught—it found a dozen errors I missed! The whole thing took it twelve minutes!'

  Ninheimer looked over the sheaf, with the neatly printed marks and symbols in the margins. He said, 'It is not as complete as you and I would have made it. We would have entered an insert on Suzuki's work on the neurological effects of low gravity.'

  'You mean his paper in Sociological ReviewsV

  'Of course.'

  'Well, you can't expect impossibilities of Easy. It can't read the literature for us.'

  'I realize mat. As a matter of fact, I have prepared the insert. I will see the machine and make certain it knows how to—uh—handle inserts.'

  'It will know.'

  'I prefer to make certain.'

  Ninheimer had to make an appointment to see Easy, and then could get nothing better than fifteen minutes in the late evening.

  But the fifteen minutes turned out to be ample. Robot E2-27 understood the matter of inserts at once.

  Ninheimer found himself uncomfortable at close quarters with the robot for the first time. Almost automatically, as though it were human, he found himself asking, 'Are you happy with your work?'

  'Most happy, Professor Ninheimer,' said Easy solemnly, the photocells that were its eyes gleaming their normal deep red.

  'You know me?'

  'From the fact that you present me with additional material to include in the galleys, it follows that you are the author. The author's name, of course, is at the head of each sheet of galley proof.'

  'I see. You make—uh—deductions, then. Tell me———'

  he couldn't resist the question—'what do you think of the book so far?'

  Easy said, 'I find it very pleasant to work with.'

  'Pleasant? That is an odd word for a—uh—a mechanism without emotion. I've been told you have no emotion.'

  'The words of your book go in accordance with my cir­cuits,' Easy explained. 'They set up little or no counter-potentials. It is in my brain paths to translate this mechani­cal fact into a word such as "pleasant." The emotional con­text is fortuitous.'

  'I see. Why do you find the book pleasant?'

  'It deals with human beings, Professor, and not with inorganic materials or mathematical symbols. Your book attempts to understand human beings and to help increase human happiness.'

  'And this is what you try to do and so my book goes in accordance with your circuits? Is that it?'

  'That is it, Professor.'

  The fifteen minutes were up. Ninheimer left and went to the university library, which was on the point of closing. He kept them open long enough to find an elementary text on robotics. He took it home with him.

  Except for occasional insertion of late material, the galleys went to Easy and from him to the publishers with little intervention from Ninheimer at first—and none at all later.

  Baker said, a little uneasily, 'It almost gives me a feeling of uselessness.'

  'It should give you a feeling of having time to begin a new project,' said Ninheimer, without looking up from the notations he was making in the current issue of Social Science Abstracts.

  'I'm just not used to it. I keep worrying about the galleys. It's silly, I know.'

  'It is.'

  'The other day I got a couple of sheets before Easy sent them off to———'

  'What!' Ninheimer looked up, scowling. The copy of Abstracts slid shut. 'Did you disturb the machine at its work?'

  'Only for a minute. Everything was all right. Oh, it changed one word. You referred to something as "criminal"; it changed the word to "reckless." It thought the second adjective fit in better with the context.'

  Ninheimer grew thoughtful. 'What did you think?'

  'You know, I agreed with it. I let it stand.'

  Ninheimer turned in his swivel-chair to face his young associate. 'See here, I wish you wouldn't do this again. If I am to use the machine, I wish the—uh—full advantage of it. If I am to use it and lose your—uh—services anyway because you supervise it when the whole point is that it requires no supervision, I gain nothing. Do you see?'

  'Yes, Dr. Ninheimer,' said Baker, subdued.

  The advance copies of Social Tensions arrived in Dr. Ninheimer's office on the eighth of May. Hejbooked through it briefly, flipping pages and pausing to read a paragraph here and there. Then he put his copies away.

  As he explained later, he forgot about it. For eight years, he had worked at it, but now, and for months in the past, other interests had engaged him while Easy had taken the load of the book off his shoulders. He did not even think to donate the usual complimentary copy to the university library. Even Baker, who had thrown himself into work and had steered clear of the department head since receiving his rebuke at their last meeting, received no copy.

  On the sixteenth of June that stage ended. Ninheimer received a phone call and stared at the image in the 'plate with surprise.

  'Speidell! Are you in town?'

  'No, sir. I'm in Cleveland.' Speidell's voice trembled with emotion.

  'Then why the call?'

  'Because I've just been looking through your new book! Ninheimer, are you mad? Have you gone insane?''

  Ninheimer stiffened. 'Is something—uh—wrong?' he asked in alarm.

  'Wrong? I refer you to page 562. What in blazes do you mean by interpreting my work as you do? Where in the paper cited do I make the claim that the criminal person­ality is nonexistent and that it is the Law-enforcement agencies that are the true criminals? Here, let me quote———'

  'Wait! Wait!' cried Ninheimer, trying to find the page. 'Let me see. Let me see… Good God!'

  'Well?'

  'Speidell, I don't see how this could have happened. I never wrote this.'

  'But that's what's printed! And that distortion isn't the worst. You look at page 690 and imagine what Ipatiev is going to do to you when he sees the hash you've made of his findings! Look, Ninheimer, the book is riddled with this sort of thing. I don't know what you were thinking of—but there's nothing to do but get the book off the market. And you'd better be prepared for extens
ive apologies at the next Association meeting!'

  'Speidell, listen to me———'

  But Speidell had flashed off with a force that had the 'plate glowing with after-images for fifteen seconds.

  It was then that Ninheimer went through the book and began marking off passages with red ink.

  He kept his temper remarkably well when he faced Easy again, but his lips were pale. He passed the book to Easy and said, 'Will you read the marked passages on pages 562, 631, 664 and 690?'

  Easy did so in four glances. 'Yes, Professor Ninheimer.'

  'This is not as I had it in the original galleys.'

  'No, sir. It is not.'

  'Did you change it to read as it now does?'

  'Yes, sir.'

  'Why?'

  'Sir, the passages as they read in your version were most uncomplimentary to certain groups of human beings. I felt it advisable to change the wording to avoid doing them harm.'

  'How dare you do such a thing?'

  'The First Law, Professor, does not let me, through any inaction, allow harm to come to human beings. Certainly, considering your reputation in the world of sociology and the wide circulation your book would receive among scholars, considerable harm would come to a number of the human beings you speak of.'

  'But do you realize the harm that will come to me now?'

  'It was necessary to choose the alternative with less harm.'

  Professor Ninheimer, shaking with fury, staggered away.

  It was clear to him that U.S. Robots would have to account to him for this.

  There was some excitement at the defendants' table, which increased as Prosecution drove the point home.

  'Then Robot EZ-27 informed you that the reason for its action was based on the First Law of Robotics?'

  'That is correct, sir.'

  'That, in effect, it had no choice?'

  'Yes, sir.'

  'It follows then that U.S. Robots designed a robot that would of necessity rewrite books to accord with its own conceptions of what was right. And yet they palmed it off as simple proofreader. Would you say that?'