Many are bothered by these possibilities, so much so that they reject them out of hand as unethical, immoral. Although I do not do so, I do sympathize with their concerns. However, I see the development of intelligent machines as both inevitable and beneficial. Where will there be benefits? In such areas as doing dangerous tasks, driving automobiles, piloting commercial vessels, in education, in medicine, and in taking over routine work. Where might there be moral and ethical concerns? Pretty much in the same list of activities. Let me explore the beneficial aspects in more detail.
Consider some of the benefits. Robots could be—and to some extent already are—used in dangerous tasks, where people’s lives are at risk. This includes such things as search-and-rescue operations, exploration, and mining. What are the problems? The major ones are likely to come from the use of robots to enhance illegal or unethical activities: robbery, murder, and terrorism.
Will robot cars replace the need for human drivers? I hope so. Every year, tens of thousands of people are killed, and hundreds of thousands seriously injured through motor vehicle accidents. Wouldn’t it be nice if automobiles were as safe as commercial aviation? Here is where automated vehicles can be a wonderful saving. Moreover, automated vehicles could drive more closely to one another, helping to reduce traffic congestion, and they could drive more efficiently, helping to solve some of the energy issues associated with driving.
Driving an automobile is deceptively simple: most of the time it takes little skill. As a result, many are lulled into a false sense of security and self-confidence. But when danger arises, it does so rapidly, and then the distracted, the semiskilled, the untrained, and those temporarily impaired by drugs, alcohol, illness, fatigue, or sleep deprivation are often incapable of reacting properly in time. Even well-trained commercial drivers have accidents: automated vehicles will not reduce all accidents and injuries, but they stand a good chance of dramatically reducing the present toll. Yes, some people truly enjoy the sport of driving, but these could be accommodated on special roads, recreational areas, and race tracks. Automation of everyday driving would lead to loss of jobs for drivers of commercial vehicles, but with a saving of life, overall.
Robot tutors have great potential for changing the way we teach. Today’s model is far too often that of a pedant lecturing at the front of the classroom, forcing students to listen to material they have no interest in, that appears irrelevant to their daily lives. Lectures and textbooks are the easiest way to teach from the point of view of the teacher, but the least effective for the learner. The most powerful learning takes place when well-motivated students get excited by a topic and then struggle with the concepts, learning how to apply them to issues they care about. Yes, struggle: learning is an active, dynamic process, and struggle is a part of it. But when students care about something the struggle is enjoyable. This is how great teaching has always taken place—not through lecturing, but through apprenticeship, coaching, and mentoring. This is how athletes learn. This is the essence of the attraction of video games, except that in games, what students learn is of little practical value. These methods are well known in the learning sciences, where they are called problem-based, inquiry-learning, or constructivist.
Here is where emotion plays its part. Students learn best when motivated, when they care. They need to be emotionally involved, to be drawn to the excitement of the topic. This is why examples, diagrams and illustrations, videos and animated illustrations are so powerful. Learning need not be a dull and dreary exercise, not even learning about what are normally considered dull and dreary topics: every topic can be made exciting, every topic excites the emotions of someone, so why not excite everyone? It is time for lessons to become alive, for history to be seen as a human struggle, for students to understand and appreciate the structure of art, music, science, and mathematics. How can these topics be made exciting? By making them relevant to the lives of each individual student. This is often most effective by having students put their skills to immediate application. Developing exciting, emotionally engaging, and intellectually effective learning experiences is truly a design challenge worthy of the best talent in the world.
Robots, machines, or computers can be of great assistance in instruction by providing the framework for motivated, problem-based learning. Computer learning systems can provide simulated worlds in which students can explore problems in science, literature, history, or the arts. Robot teachers can make it easy to search the world’s libraries and knowledge bases. Human teachers will no longer have to lecture, but instead can spend their time as coaches and mentors, helping to teach not only the topic, but also how best to learn, so that the students will maintain their curiosity through life, as well as the ability to teach themselves when necessary. Human teachers are still essential, but they can play a different, much more supportive and constructive role than they do today.
Moreover, although I believe strongly that we could develop efficient robot tutors, perhaps as effective as Stephenson’s The Young Lady’s Illustrated Primer (see page 171), we would not have to abandon human teachers: automated tutors—whether books, machines, or robots—should act as supplements to human instruction. Even Stephenson writes in his novel that his star pupil knew nothing of the real world and of real people because she had spent far too much time locked up in the fantasy world of the Primer.
Robots in medicine? Yes, they could be used in all its aspects. In medicine, however, as in many other activities, I foresee this as a partnership, where well-trained human medical personnel work with specialized robotic assistants to increase the quality and reliability of care.
Laser surgery on eyes is now close to complete machine control, and any activity where great precision is required is a candidate for machine operation. Machine diagnosis is trickier, and I suspect that skilled physicians will always be involved, but that they will be aided by dynamic, intelligent machines that can assess a large database of prior cases, medical records, medical knowledge, and pharmaceutical information. This assistance is already required, as the amount of information and the rapid addition of new information becomes overwhelming to practicing physicians. Moreover, as we get better diagnostic tools—more efficient analyses of body fluids and physiological records, DNA analyses, and various body scans—where some of the information will be routinely collected and sent from a patient’s home or even place of work to the medical office, only a machine could keep up with the information. People are excellent at synthesis, at dynamic, creative decisions, at seeing the whole, global picture, whereas machines are superb at rapid search through large numbers of cases and information files, without being subject to the biases that accompany human memory. The team of trained physician and robotic assistant would be far superior to either working alone.
One common fear, of course, is that robots will take over many routine jobs from people, therefore leading to great unemployment and turmoil. Yes, more and more machines and robots will take over jobs, not only of lower-skilled workers, but increasingly of much routine work of all kinds, including some management. Throughout history, each new wave of technology has displaced workers, but the total result has been increased life span and quality of living for everyone, including, in the end, increased jobs—although of a different nature than before. In transitional periods, however, people are displaced and unemployed, for the new jobs that result often require skills very distant from those of the people who have been displaced. This is a major social problem that must be addressed.
In the past, most of the jobs replaced by automation have been lowlevel jobs, jobs that did not require much skill or education to perform. In the future, however, robots are apt to replace some highly skilled jobs. Will film actors be replaced by computer-generated characters that sound and act just as realistic, but are much more under the control of the director? Will robot athletes compete, if not with humans, then perhaps in their own leagues—but thereby leading to the demise of human leagues? Such a situation might very well hap
pen with chess tournaments and leagues, now that computer chess players can beat even the best human players. What about jobs such as accounting, bookkeeping, drafting, stock keeping, or even simple management jobs? Will these be replaced? Yes, all this is possible; some of it has already started. Robot musicians? The list of potential activities is large, along with the dangers of social upheaval.
When robots are used for activities such as space exploration, dangerous coal mining, or search-and-rescue missions, or even when they do simple things around the house, such as vacuum cleaning and other chores, there is not apt to be much resistance. But when they start taking over large numbers of jobs or displacing large amounts of people from routine activities, then this does become a legitimate concern, one that raises serious issues for society.
I believe that we should welcome machines that eliminate the dreary tedium of many jobs—the dull shuffling of paperwork probably being even more demeaning than many of the low-paid, routine service jobs. This welcome, of course, assumes that machines will free people to engage in more creative activities, where they can apply their abilities both more pleasurably and effectively.
I have visited many parts of the world where poverty, continual hunger and starvation, and high death rates have made me doubt the benefits of today’s systems. I have seen silk factories in India where young girls are locked into buildings, forced to weave from early morning till evening, locked in so that they cannot leave—or even escape the building if there is fire—without someone from the outside unlocking the doors. My study of history has taught me that such inequity, brutality, and callous treatment of so many is not unusual, and long predates the development of modern technology.
Yes, I see the downside of the deployment of intelligent machines and robots, but I also see the downside of no deployment. Call me an optimist, if you wish, but I believe that in the end, the human ingenuity that we show in creating these powerful devices will also serve us in ways to create more enriching, more enlightened activities for all of us. Optimism does not blind me to the inequities and problems of today’s life: optimism reflects my belief that we can overcome them in the future. Yes, we still have poverty, starvation, political inequity, and wars, but these result more from the evils of people than from our technologies. I do not see why the introduction of smart, emotional robots and machines will change this situation, either for the worse or for the better. To change evil, we must confront it directly. It is a social, political, and human problem, not a technological one. This, of course, does not minimize this problem nor does it absolve us from working toward a solution. But the solution must be social and political, not technological.
The story becomes even more complex if I expand the view beyond the short-term horizon. At some point, robots and other machines are apt to become truly autonomous. This is a long time away, perhaps centuries, but it will happen. Then, there will indeed be major disruptions of life when much or all human work can be done by robots: farming, mining, manufacturing, distribution, and sales. Education and medicine. Even many aspects of art, music, literature, and entertainment. Robots may manufacture themselves. At that point, the relationship between natural animals and robots becomes exceedingly complex. The complexity will be amplified because many humans will actually be cyborgs—part human, part machine. Artificial implants already exist, mostly as medical prostheses; but some people are talking about having them implanted on demand, the better to enhance natural capabilities. Strength, athletic ability, sensory capability, memory, and decision making could all be aided by implanted, electronic, chemical, mechanical, biological, or nanotechnology devices. Steroids are used by athletes to enhance their existing strength, and laser treatment of the cornea has been done by some athletes and pilots to enhance normal acuity. The artificial lenses in my eyes—implanted after cataract removal—have provided me with far better vision than I have ever had before, with the sole problem being that my eyes cannot change their focus. But someday, artificial lenses will be able to focus, probably even better than natural ones, perhaps providing telescopic in addition to normal vision. When this happens, even people who do not have cataracts might wish to have their normal lenses replaced by these more effective ones. Even more striking artificial enhancement is possible. Such possibilities raise complex ethical issues, but these truly go beyond the boundaries of this book.
But this book does focus upon emotions and their role in the development of artificial devices and the way that human beings emotionally attach themselves to their belongings, their pets, and to one other. Robots might act like all of these. At first, robots will be belongings, but ones with clear personal attachment, for if a robot is with you for a large part of your life, able to interact, to remind you of your experiences, to give advice, or even just comic relief, there will be strong emotional attachments. Even today’s robot pets, crude though they may be, have already evoked strong emotions among their owners. In the decades to come, robot pets may take on all the attributes of real pets and, in the minds of many people, be superior. Today people abuse and abandon their pets. Many communities have bands of stray cats or abandoned dogs scavenging. Might the same happen with robotic pets? Who is legally responsible for their care and maintenance? What if a robot pet injures someone? Who is legally responsible? The robot? The owners? The designer or manufacturer? With real pets, the owner is responsible.
And finally, what happens when robots act as independent, sentient beings, with their own hopes, dreams, and aspirations? Will something akin to Asimov’s Laws of Robotics be necessary? Will they be sufficient? If robot pets can cause damage, what might an autonomous robot do; and if a robot causes damage, injury, or death, who is to blame, and what is the recourse? Asimov concluded in his novel I, Robot, that robots will indeed take over, that mankind will lose its own say in its future. Science fiction? Yes, but all future possibilities are fiction before they are fact.
We are in a new era. Machines are already smart, and they are getting smarter. They are developing motor skills, and soon they will have affect and emotion. The positive impact will be enormous. The negative consequences will also be significant. This is how it is with all technology: it is a two-edged sword always combining potential benefits with potential deficits.
EPILOGUE
We Are All Designers
I TRIED AN EXPERIMENT. I posted a request to some internet discussion groups for examples of products and web sites that they loved, hated, or had a love/hate relationship with. I received around 150 email responses, many passionate, and each listing several items. The responses were highly biased toward technology, not surprisingly, because this is the area in which most of the respondents worked; but technology did not receive high marks.
One of the problems with such a survey is the “too obvious to notice” effect, as reflected by the old folk tale that a fish is the last to notice water. Thus, if you ask people to describe what they see in the room in which they are sitting, they are apt to leave out the obvious: floor, walls, ceiling, and sometimes even windows and doors. People may not have reported what they truly liked because that might have been too close to them, too enmeshed in their lives. Similarly, they might have missed the disliked things because they were absent. Still, I found the responses interesting. Here are three examples:Global chef’s knives—beautiful, functional and simple. They are delightful to hold and use. I keep mine under my pillow (youch! just kidding).
The pièce de résistance is my watch. A George Jensen: sterling, large mirror face with two arms but no markings for numbers, the arm band is incomplete, only covers 3/4 of your wrist. Out of the ordinary, beautiful. (The design is in the Museum of Modern Art.) P.S. I stared at it, at least six years in Paris before I bought it.
My VW Bug: love it—it’s simple, utilitarian, gets great gas mileage, small enough to park just about anywhere and just plain fun to drive. But I can’t get past that stupid seat-lift-handle thing—it drives me bonkers. (The lift-the-seat handle on the front seat
s—they’re in the “wrong” spot. Not one person has ever “gotten it right.”)
Love it, hate it, indifferent to it. Our interaction with our everyday things reflects the three levels of design in very different ways. Loved objects ranged the gamut of all possible combinations of the three forms of design. Many an item was enjoyed solely for the visceral impact of its appearance:After plunking down $400 for an iPod I almost wouldn’t have cared about the product after having unwrapped the packaging, it was that nice. [The iPod is Apple Computer’s music player.]
I bought a VW Passat because the controls inside the car were pleasurable to use and look at. (Get in one at night—the dashboard lights are blue and red-orange.) It makes driving more fun.
Remember the person in chapter 3 who bought water simply because the bottle looked so great? That response certainly belongs in this category:I remember deciding to buy Apollinaris, a German mineral water, simply because I thought it would look so good on my shelves. As it turned out, it was a very good water. But I think I could have bought it even though it was not all that great.
Many products were loved for their behavioral design alone—that is, their function and utility, usability and understanding, and physical feel:I like my OXO vegetable peeler, too. It handles eggplant, broccoli stems and anything else I throw at it. They make those nice comfy handles.
Lie-Nielsen hand planes: I can plane tiger maple and produce a smooth, glassy, surface where most planes would tear out chunks of wood.
Can opener: You may recall Victor Papanek’s short book How Things Don’t Work. In it he mentions a can opener. I finally found it a few years back—it’s been reproduced by Kuhn Rikon as their LidLifter Can Opener. In brief, it opens the can by splitting the seam, rather than cutting through the top. Lots of reasons why that’s a good thing, but it’s an appliance I actually look forward to using. Hand operated, needs little cleaning, fits my hand, does its job, stores in a drawer, easily accessible. A dutiful servant, as a kitchen appliance should be.
Emotional Design Page 21