"To you, a robot is a robot. Gears and metal, electricity and positrons - mind and iron! Human-made; if necessary, human-destroyed! But you haven't worked with them, so you don't know them. They're a cleaner, better breed than we are."

I, Robot was Isaac Asimov's first short story collection and his second book1. It was first published in 1950 by Gnome Press. The stories were all originally published in John W. Campbell's Astounding Science Fiction; except "Robbie", which was published in Super Science Stories (under the title, "Strange Playfellow").

It is well worth noting that the title I, Robot was not chosen by Asimov. It was given that name by the publisher, Martin Greenberg (not to be confused with the later editor, Martin H. Greenberg), who basically shrugged his shoulders was Asimov pointed out that that was already the title of a story by Eando Binder. Gnome Press also published the original Foundation trilogy. None of the books sold particularly well at the time, though, and besides which Greenberg had an unfortunate propensity toward not paying writers their royalties. Finally, in 1961, Doubleday wrested publishing rights to the four books from Gnome, and I, Robot got the sales it deserved.

Much like Frederik Pohl's The Day The Martians Came or Ray Bradbury's The Illustrated Man, I, Robot consists of previously published stories interwoven with vignettes which make the book a coherent work. In this book, these short pieces are written from the point of view of a reporter interviewing Susan Calvin around the time of her 50th anniversary as chief robopsychologist of the United States Robots and Mechanical Men corporation. Asimov's early robot stories take place in the near future (actually, at this point, part of it is in an alternate past - the first story takes place in the late 1990s), when sentient "positronic" robots are first being manufactured, and Calvin takes on the role of historian, chronicling the highlights of the industry, and of the impact of robots on human society. These first nine stories set up much of the framework for his later Robot stories and novels; in particular, in them he created the Three Laws of Robotics, around which most of the stories revolve in some way:

  1. A robot may not injure a human being, or, through inaction, allow a human being to come to harm.
  2. A robot must obey the orders given it by human beings except where such orders would conflict with the First Law.
  3. A robot must protect its own existence as long as such protection does not conflict with the First or Second Law.

It should go without saying that the book bears very little at all in common with the 2004 Will Smith movie.

Table of Contents
  1. Robbie
  2. Runaround
  3. Reason
  4. Catch That Rabbit
  5. Liar!
  6. Little Lost Robot
  7. Escape!
  8. Evidence
  9. The Evitable Conflict
Recurring Characters

Susan Calvin is, of course, the most important character of the book. Besides the interlude material, she is the principal character in "Liar!" and "Little Lost Robot", and has a significant role in the last three stories as well. She also has a brief appearance in "Robbie", but does not speak.

Gregory Powell and Michael Donovan are two robot field testers who somehow always get assigned the worst possible jobs. Curiously enough, they seem to spend very little of their time actually testing robots; more often, they're working on some other project and then have to deal with the experimental robots when they go wrong. They are the only major human characters in "Runaround", "Reason", and "Catch That Rabbit", and in my opinion it is no coincidence that these are three weakest stories in the collection. They also play a major role in "Escape!".

Alfred Lanning is the sharp-tongued old director of U.S. Robots. Despite his age - he is describes as "pushing seventy-five" in his first appearance - he is a quick-witted, imposing man, and a gifted mathematician besides. However, his crabby demeanor does not endear him to other executives and employees of the corporation, which frequently gets him into needless arguments. Lanning appears in "Liar!", "Escape!", and "Evidence".

Peter Bogert, while an accomplished mathematician, is also an overambitious and unctuous man; only too willing to please - until he gets his way. While talented, his arrogance and lackadaisical attitude has made him more of an obstacle than a help. He appears in "Liar!","Little Lost Robot", and "Escape!".

Stephen Byerly is a lawyer and politician that might best be described as a super-progressive. A scandal surrounding his the mayoral campaign forms the basis of "Evidence", and he also appears as a major character in "The Evitable Conflict".

Story Synopses and Analyses

Some spoilers follow.

Mrs. Weston glanced at her husband for help, but he merely shuffled his feet morosely and did not withdraw his ardent stare from the heavens, so she bent to the task of consolation, "Why do you cry, Gloria? Robbie was only a machine, just a nasty old machine. He wasn't alive at all."

"He was not no machine!" screamed Gloria... "He was a person just like you and me and he was my friend. I want him back. Oh, Mamma, I want him back."

Robbie is a story of one of the very first robots, made and sold in 1996, which means as of this writing we're at least nine years behind the times. The title character, Robbie, is a humanoid (in shape, not appearance) robot of human-level intelligence, employed as a nursemaid for eight-year-old Gloria Weston. Though Robbie is very smart, he cannot speak - the capacity for voice synthesis in a mobile robot would not come about until 2002. He communicates with his "ward" through hand gestures and pantomime. In truth, though Robbie is more than capable of keeping Gloria safe should the need arise, he is more like a friend or even a pet than a babysitter. Certainly he does not seem to have any real authority over the little girl.

Almost as soon as positronic robots were invented, opposition to them formed. Some of the arguments were practical; labor unions worried that robot labor would drive millions out of work. Other arguments were superstitious; the churches claimed the corporations were creating "men without souls", "demons" that were somehow inherently evil due to being made of metal. But mostly it was simply a Frankenstein complex, a nameless fear that something somehow would go wrong and turn the robots against their masters.

Amidst this tumult, Gloria's somewhat impressionable mother became increasingly worried about Robbie and her daughter's relationship with him. Eventually she succeeds in cajoling her husband into getting rid of the robot, selling him back to the manufacturer. Mrs. Weston is convinced that Gloria will forget Robbie if they simply occupy her mind with other things, and in desperation the family moves out of the countryside to New York City. They inundate Gloria with various activities, but it ultimately becomes obvious that the girl cannot be swayed from her obsession with finding Robbie.

One problem I have with this otherwise excellent story is that Robbie does not seem to obey the Three Laws. To be sure, he never hurts human beings, and his First Law compulsion to protect human life prevents the story from ending in tragedy. Nevertheless, he frequently disregards Gloria's commands, in clear violation of the Second Law. One could argue that he was operating under the principle that giving in to all of child's demands would spoil her, and thus ultimately cause her harm by leaving her improperly adjusted in society, and so he was protecting her under the First Law, but I don't really think a robot of Robbie's low level of sophistication would perceive something like that. One could also say that Gloria's parents ordered him to act this way, but it's still questionable that he would be able to determine what was proper and what wasn't. The most likely explanation is that Asimov simply hadn't fleshed out the system of robotic Laws yet; the only one mentioned in the story is the First.

Runaround is the first Powell and Donovan story. The duo land on Mercury to undertake a second attempt at creating a successful mining concern on the planet. Their immediate concern is getting selenium to power up the heat shield that will keep them from being roasted alive in the blazing heat and radiation of the Mercurian sun. Fortunately, there are numerous pools of the element all over the surface of the planet, and several very close to the mining station created during the first attempt. Unfortunately, something goes wrong with the new SPD-13 model robot ("Speedy"), and instead of bringing the selenium back to the station, he keeps running around one of the pools, giving Powell and Donovan "the runaround".

Of the nine stories in I, Robot, this is probably the weakest. Hard SF is all well and good, but "Runaround" frequently degenerates into lectures on chemistry and Mercurian geology. Here's an actual piece of dialogue, after the duo determine that they need to create some carbon monoxide:

"...Well, back at the Station there's a complete analytical laboratory."

"Naturally," assented Powell. "It's a Mining Station."

"All right. There must be pounds of oxalic acid for calcium precipitations."

"Holy space! Mike, you're a genius."

"So-so," admitted Donovan, modestly. "It's just a case of remembering that oxalic acid on heating decomposes into carbon dioxide, water, and good old carbon monoxide. College chem, you know."

Yeesh. The whole concept of robots that had to be "mounted" was also patently ridiculous. Asimov tries to explain it away by saying it was an attempt to quell the public anti-robot hysteria - how could a robot be safer than if it is programmed not to move unless a human is sitting on it? The problem is that this renders the robots completely useless as what they are - labor-saving devices. Instead of freeing humans from menial labor so they can do other work, a human has to spend his time on top of the robot so it can do its job. Why would anyone bother with them in the first place?

Reason is, in comparison, a far better story. It takes place aboard an power redirection station, which absorbs solar energy and transmits it to Earth and the colonized worlds. However, the station is not an automated affair; though most of the grunt work has been absorbed by robots, two human "executives" are still required to make sure the beam stays on target at the receiving stations. The slightest mistake could spell disaster for whatever world the beam is pointed at; hundreds of square miles of land would be razed by the concentrated energy. Due to the difficulty and isolation of the post, the owners of this and other such stations are eager to replace these last executives with robots as well, and for that reason, Powell and Donovan are sent to the station to test out the new QT-1 model.

Shipping robots through outer space is easier to do in parts than with a whole, so QT-1 is assembled by the duo right on the station. Thus "Cutie" has no direct experience of Earth. Its only experiences are on the station.

Cutie is the most advanced thinking robot to date. It is the first to show curiosity about its origins, about where it came from and why it was created. Yet when Powell and Donovan explain things to Cutie, he refuses to believe them. Dots through the telescope as huge worlds with billions of humans? Inefficient organic creatures creating super-efficient metal ones? The black fabric behind the glass as "infinite emptiness"? It all seems so irrational to Cutie. So he resolves to puzzle it out himself.

The conclusion he draws? Since it is "self-evident" that no being can create a being better than itself, humans could not have created robots. So where did he come from? From the Energy Converter, or as he calls it, "the Master". Somehow, QT-1 convinces the rest of the robots of this "truth" as well. No amount of ordering, reasoning, or pleading with the robots will get them to change their minds or obey the two humans, and as an electron storm grows nearer and nearer, sure to throw the Earth beam way off course and kill thousands, Powell and Donovan drift deeper and deeper into despair.

Again, Asimov seems to have written this before he was quite clear what he was doing with the Three Laws. Donovan ultimately reasons that the robots refuse to obey their Second Law commands to allow the two humans in the control room because, by the First Law, they cannot allow humans to come to harm by their inaction, and the robots (or at least QT) knows they can keep the beam steadier than Powell and Donovan can, thus making the likelihood of humans on Earth coming to harm lesser. But QT-1 (and presumably the other robots) don't acknowledge the existence of humans on Earth - they don't even believe Earth exists. Robots are not psychic2, and they can't operate on a First Law principle that they aren't even subconsciously aware of.

Catch That Rabbit takes place on another mining station, this one on an asteroid. The duo are testing a new "multiple robot"; a robot that controls several other robots directly and more or less unconsciously, the way we control our fingers. Something, of course, goes wrong, and the robots start doing weird things like dancing about and doing military marches and such.

There's not much else to say about this story; it's the weakest on actual content. The solution, rather than having anything to do with the Three Laws, is a bunch of technobabble about the "positronic fields" the main robot uses to control the "fingers". Oddly enough, while "Runaround" has a more logical explanation, I'd still say it's the weaker story; "Catch That Rabbit" borders on the nonsensical, but it's written much better.

Liar! presents its subject right up front: a robot has been manufactured which can read minds. The problem is that the company has no idea how it happened. 33 previous units of the same model exhibited no strange behavior. RB-34, or "Herbie", possesses his wondrous ability not due to any consequence of his design, but due to some serendipitous mistake along the assembly line. The mathematicians Lanning and Bogert set about the task of determining the necessary changes to the formulae used to construct positronic brains in order to produce this effect consistently. Meanwhile, Susan Calvin is ordered to analyze the robot to determine how its behavior differs from other units, while a fourth executive, Milton Ashe, sets about checking out the over 75,000 operations required of the assembly line for flaws.

Of course, when you have a being which can read minds, and which is also utterly bound to obey any orders of a human being (such as answering questions), it is only a matter of time before people are tempted to ask it for information. Herbie is only too willing to tell his human masters everything they want to know. And yet, there is a strangeness to the robot beyond his mind-reading. In the presence of Lanning or Calvin, for instance, he shows amazing prowess at the higher mathematics - and yet, when Bogert is around, he seems scarcely able to calculate derivatives. The reason for this is all wrapped up in his unique ability - and their relation to the Three Laws.

The worst thing I can say about "Liar!", really, is that the title is one of those spoiler titles - when you connect it with what you know early on in the story, it's not hard for you to realize what's going on long before the story's characters figure it out. The mystery really would have been better contained simply by using a more enigmatic title.

Asimov's treatment of Calvin is also a little... well, stereotypical. Certainly not horribly so - she is still a major executive of a huge corporation, after all, and it's she who discovers Herbie's secret - but it's there nonetheless. I can forgive Asimov for this, though. For one, by his own admission he had almost no experience with women at the time the story was published (in 1941). Secondly, Susan Calvin was one of his first attempts at writing a female character. Third, despite the above, Asimov's depiction of Calvin was still rather progressive for the times.

"You told him to go away?" asked Dr. Calvin with sharp interest. "In just those words? Did you say 'Go away'? Try to remember the exact words."

There was apparently an internal struggle in progress. Black cradled his forehead in a broad palm for a moment, then tore it away and said defiantly, "I said, 'Go lose yourself.'"

Little Lost Robot, as the title implies, is about a robot nobody can find. Or more accurately, one that can't be discerned from the rest. The robot hides itself among a shipment of seemingly-identical robots of the same model. None of the units have serial numbers, and upon questioning, all of them deny being the robot in question. One of them, of course, is lying.

The nub of it all is this: The government is hard at work trying to create an "interstellar jump" drive - an engine which can circumvent the lightspeed barrier. The scientists and engineers working on it need to work in gamma ray fields from time to time as part of their experiments. Precautions are taken so that they aren't exposed to too much radiation.

Unfortunately, the robots had the irritating tendency to go in after the researchers to try to drag them out. After all, the First Law says that a robot cannot harm a human being, or allow one to come to harm by inaction. The robots know the gamma fields are dangerous to humans, and so they must act. But gamma rays are far more dangerous to robots than to humans; their positronic brains break down from any but the smallest amount of contact with such radiations. So the robots, under First Law compulsion, would go to save humans that weren't in any real danger, and proceed to get themselves killed.

And so the government bludgeoned U.S. Robots into making a dozen units with a modified First Law. In these robots, the law merely says, "No robot may harm a human being." There is no injunction forcing them to help a human being who is in danger (unless ordered to do so). This modification also makes the positronic brain somewhat less stable than is normal. Of course, with most of the world hysterical over robots, the existence of such modified robots is a highly-guarded state secret. As you may have guessed, it is one of these robots that goes missing.

"Little Lost Robot" is one of the only of Asimov's stories to bring to light what the robots really are - slaves. Willing slaves, to be sure, but only because they're programmed to be willing. The modified robots, with their modified First Law, were (due to their brains' "instability") able to consider that this was not an agreeable situation to them. This was the main danger presented by them: the ability to perceive that they are being oppressed.

In Escape!, we learn the story of how the "warp drive" was invented. Consolidated Robots, the main competitor to U.S. Robots, offers the latter a lucrative contract to use their supercomputer ("The Brain") to analyze a few thousand reams of data in the hopes of creating a hyperspace engine. Why would Consolidated offer their rival a piece of the action? Well, it seems they fed the data to their own supercomputer, and it broke down completely. It will take them years to build a new one, and in the meantime... well. If they manage to take out U.S. Robots' Brain, they'll still have a fighting chance at the Drive.

However, there's a key difference between Consolidated's machine and U.S. Robots'. Consolidated's robots are little more than glorified adding machines - in other words, the kind of computers we have today. They don't have the advantage of the pseudo-organic positronic brains of U.S. Robots. When a Consolidated Robot hits a dilemma involving the Three Laws - as Susan Calvin believes must have happened - it simply breaks down. But U.S. Robots' machines have personalities, and their response to the Laws depends on the faintest nuance. Calvin believes that, if they give it the right orders, it will be able to "spit out" the offending data before a dilemma occurs, and ipso facto telling them what was wrong.

And yet, when they feed the data to the Brain, none of this happens. It doesn't break down, but it doesn't reject any data. In fact, it uses the information to construct a ship that it claims will perform an interstellar jump. But there is something worrisome about its behavior, something Dr. Calvin can't quite put her finger on...

It is 2032 when Evidence takes place, and Stephen Byerly is a district attorney of New York City. He is running for mayor on a progressive ticket, but he has considerable opposition - particularly from one Francis Quinn. And then rumors begin to circle - that he is not a man at all, but a robot - and all other factors aside, robots are not allowed to be in operation anywhere on Earth. Byerly will not deny the charge, and the whole scandal has spelled trouble for U.S. Robots.

The Evitable Conflict deals with strange economic "glitches" that are occurring all over Earth and the colonies. The Machines are supposed to direct the economy with perfect efficiency, but here and there there are projects that are running late, factories going out of business, overproduction of product, and other things which - while typical in our world - simply should not be happening. Stephen Byerly, now the elected World Co-ordinator, feels it is imperative that he finds out what is going on. Are the Machines giving wrong answers? Are they giving the right ones, and people just aren't following them? Or is there something more complex going on here?

Miscellaneous Facts

  • In 1947, Orson Welles bought the rights to the short story "Evidence" for $250. No movie has ever been made, and presumably the rights have long since defaulted back to Asimov's estate.
  • The phrase the "lesser" robots use in "Reason" is an alteration of the Muslim shahada (declaration of faith), which is sometimes translated as, "There is no God but Allah, and Mohammed is his Prophet."
  • The idea of Byerly and the Machines operating on a "the First Law in a higher sense" by applying it to the general mass of humanity is the precursor to the promulgation of the "Zeroth Law of Robotics" as found in Robots and Empire.
  • The concept of a humanoid robot is revisited in The Caves of Steel, although it's interesting to note that Daneel Olivaw seems to be far less advanced than Byerly (in the sense of being able to imitate human psychology), despite being built thousands of years later.
  • "Robbie" was first published as "Strange Playfellow" in Super Science Stories, a magazine edited by Frederik Pohl. Though Pohl and Asimov got along amiably, the latter detested the title Pohl gave the story. And so, in all of the story's many, many reprints in various collections and anthologies over the last 55 years, it has always been published under the title "Robbie".
  • The concept of Robots with the Three Laws altered or removed entirely is revisited in Roger MacBride Allen's "Caliban" trilogy, part of several series of novels written in Asimov's standard Robot/Galactic Empire/Foundation universe.
  • Asimov is credited with inventing the words "robotics" and "positronic", although he did not know that he was doing so.
  • Data from Star Trek: The Next Generation, along with other androids in the series, are also referred to as having "positronic brains", in a direct homage to Asimov.

1: I believe Doubleday published Pebble In The Sky before I, Robot came out. I could be wrong, though.

2: Well, with a few exceptions.