In a way, we even guided our own evolution, making ourselves more social beings, more capable of surviving in crowded cities, or in the grip of a fearfully complex technology. (Not that we fit in very well, but we’ve only had a short time in which to evolve these characteristics.)

Now we are beginning to be capable of genetic engineering, and our direction of evolution may become more precise and efficient (if we can make up our minds as to the particular direction in which it will be safe to proceed).

That brings us to robots, which represent what is perhaps a peculiar middle-ground between technology and life.

The robots I have pictured in my early robot stories were machines. However intelligent they seemed, they were as helpless in the grip of technology as a wheelbarrow was. They were devices that could not reproduce themselves and that, therefore, could not engage in non-directed evolution. If an improved robot was desired, a different robot, a more specialized robot, a more versatile robot, such a thing would have to be constructed by human designers.

Sure enough, as I continued to write my stories, robots did advance, grow more complicated, more intelligent, more capable-but their evolution remained directed.

What about the robotic brains? As they approached the human brain in character, might they not eventually take matters into their own hands? The brains of my robots, however, are tied tightly to the Three Laws of Robotics, and that limits them as human brains are not.

But let’s think again. Evolution is a matter of generations, of numerous individuals, each one slightly different from all the others, coming and going. A single organism in a single lifetime does not evolve in the biological sense. An individual chimpanzee does not become a human being, or even make any step, however small, toward becoming a human being in the course of its own lifetime.

If an individual organism cannot evolve by itself, it can learn, and the more complex the brain, the more efficiently and radically it can learn. Learning is a form of change, if not biologically, then at least culturally. This point does not have to be belabored in connection with human beings, but what about robots?

I reached a turning point in my own robot stories with the appearance of R. Daneel Olivaw in The Caves of Steel and of R. Giskard Reventlov in Robots and Empire. Daneel was a humaniform robot, indistinguishable from human beings if you don’t count the fact that it was far superior to human beings in a moral sense. Giskard was metallic but possessed the power of adjusting human emotions.

Each was sufficiently complex to be capable of learning, despite the weight of the Three Laws of Robotics. In Robots and Empire, Daneel and Giskard learned friendship for each other. They also labored with the concept of working for the good of humanity as something superior to the task of working for the good of individual human beings, thus groping toward what I called the “Zeroth Law of Robotics.”

In a way, robots can even offer mental complexities far beyond those in human beings. What if the “wiring” of a robot brain is replaced with another set but imperfectly so, so that a robot is aware of two sets of impressions-a kind of robotic schizophrenia? What if a robot originally intended for a particular society is forced to perform its functions in an entirely different society? How does its brain react to that? (This volume of the Robot City series involves questions of this nature.)

Can the undirected nature of robot evolution also become directed? For instance, suppose it is the task of robots to form other robots and, in particular, to design the brain patterns of other robots. This would be the robotic equivalent of genetic engineering, and robots in this way could direct their own evolution.

Or if you had humaniform robots like Daneel, and divided them into male and female with the ability of self-propagation, human fashion, a form of biological evolution might result-but then the distinction between robots and human beings would tend to disappear, and with it the possibility of meaningful robot stories.

They had named the starship the Wild Goose Chase, for when they’d left home in it some of them had doubted that the trip would be of any value. Now the ship once again orbited its world of origin, and its passengers still wondered whether they had accomplished anything useful.

They had accomplished plenty; no one disagreed about that. During their travels they had transformed one of Dr. Avery’s mutable robot cities into a toy for intelligent aliens, had reprogrammed another robot city to serve an emerging civilization on yet another alien world, had formulated a set of rules describing the motivations behind human behavior, had nearly found the mother to four of the group’s members, and had ended the career of the alien pirate who had dogged their steps for years. All the same, the operative word was “useful,” and not one of their actions received the unanimous approval of the entire crew.

None of them supposed that turning a city into a toy was anything other than an irritating lesson in futility. Derec and Ariel also had grave reservations about leaving the other robot city in the hands of the pre-technological Kin. None of the human complement-nor even Wolruf, their alien companion-cared a bit for the robots’ “Laws of Humanics,” and though Derec was excited at the prospect of finding his mother, his father harbored a contrary emotion, and besides that, they had lost her trail.

Even removing the pirate Aranimas from the picture was only a qualified success, for though they hadn’t killed him, the moral implications inherent in their method of dealing with him had driven three of the robots into the positronic equivalent of catatonia.

It was high time to go home and think about things for a while.

Home in this case meant the original Robot City, an entire planet covered with Dr. Avery’s mutable, ever-changing cybernetic metropolis. At least it had been covered in city when they left. Now, however, from their vantage point in close orbit, it looked like a newly terraformed planet still waiting for settlers.

Three humans, one alien, and a robot crowded into the starship’s control cabin to watch it drift by in the viewscreen. They were a motley-looking group by anyone’s standards. The alien, Wolruf, occupied the pilot’s chair, the demands of her canine body warping the chair into a configuration a human would have considered uncomfortable at the very least. Her brown and gold fur had been carefully brushed, but she wore no clothing or ornamentation over it.

To her right stood Derec, a thin, narrow-faced, blond-haired young man who carried the impatient look common to explorers. His clothing was utilitarian: loose pants of soft fabric suitable for anything from Yoga exercises to wiping up oil spills while dismantling machinery, capped by a plain pullover shirt of the same material, both in light blue. Snuggled close to his right stood Ariel, equally thin-though in a softer sort of way-dark-haired, and not as transparently impatient as her companion. It was obvious she had spent more time on her wardrobe than he. She, too, wore pants and a blouse, but her blouse clung where it was supposed to cling, hung loose where loose suited her figure better, exposed enough skin at neck and waist to suggest but not to provoke, and together the pale yellow and brown hues of blouse and pants provided a splash of color to offset Derec’s uniformity.

On the other side of Wolruf stood Dr. Avery.He was an older version of Derec: shorter, rounder, grayer, moustached, his face not yet wrinkled but showing the effects of time and much experience. He wore his usual baggy trousers, white shirt with ruffled collar, and oversized coat today, as most days, in gray. His expression was one of puzzlement shading over into concern.