Thinking About a Robot-Based Future

by Arnold Brown, Chairman,
and Cynthia Mooney, Senior Analyst,
Weiner, Edrich, Brown, Inc.

In coming years some of these slaves will even be living systems, made out of DNA, that will perform tasks deemed too hazardous for humans, such as cleaning up toxic waste. Others, deliberately made to resemble humans, will be companions and teachers of children. Some will even be chimeras made up partly of human cells. These will increasingly generate controversies about the extent of their humanity, if any, and what that might entitle them to. And, increasingly, they will be self-reproducing.

The capabilities of computers have expanded beyond the merely physical for some time now. Computer scientists and software programmers are increasingly outfitting the machines with emotional qualities, both in terms of conveying emotions and detecting them. Companies are already making use of computer software that detects when a person on hold is getting annoyed. Amtrak has an avatar named Julie that takes reservations, all the while putting human-like emotions into "her" voice. A group of computer scientists in South Korea is focusing on the essence of robots, rather than just their functionality. The researchers have developed software that gives robots the ability to feel emotions such as happiness, sadness, anger and fear, along with the abilities to reason and desire. Scientists at the University of Reading in England have developed the Perspex method of writing computer programs that enables computers to recover from illness and injury, much as humans do. In other words, computers that lose data and suffer damages would continue to develop their capabilities despite, or because of, the rigors of living in the world.

There are now three relatively new skills that are becoming necessary for computers to have: sensing, reasoning and communicating. These skills make possible what is known as attentive or considerate computing, enabling computers to interpret human emotional states based upon a variety of human behaviors and actions. One particular software program ranks the urgency of voice mail messages based on the volume, pitch and rate of speech of the person leaving the message. Attentive computers can scrutinize how much attention a worker is devoting to a particular task or detect when computer users have reached their cognitive limits. Microsoft has developed the Attentional User Interface, which enables computers to detect when a user is supposedly available for communication or when he or she is in a "state of flow."

It may appear that fitting computers with human-like emotions and human-like emotional skills will make them better suited to adapt to human needs. But, it is quite possible that what we will find is that humans begin adapting their emotions to the limitations of their computers. This reverse adaptation has in many ways taken place with regard to the intelligence capabilities of computers. Consider, for example, the various accounting and bookkeeping programs, such as QuickBook, that are available. Many of us have changed the ways in which we keep track of our finances, or generate reports, in order to adapt to the parameters of the computer's software. We do the same when confronted with automated help lines -- the menu may not list exactly what we are looking for, so we choose the next best option, hoping the answer to our question will be forthcoming. The same is true for on-line help options; the pull-down menu forces us to choose what the computer wants us to be searching for, not what we are looking for.

Questions are already being raised in literature, film and TV shows about the present and future relationship between humans and robots. Years ago the movie "Spartacus" showed the causes and consequences of revolt by human slaves in ancient Rome. Recent movies, such as "Blade Runner", "I, Robot" and the "Terminator" films, depict similar -- and in many ways more frightening -- revolts by non-human slaves.

Japan is actually in the forefront of robot development. Their focus is on making the robots more human-like, so they are making robots that have a sense of touch, as part of making them more acceptable for use as pets, playmates and social workers. Robots are also being made to understand not only human speech but also gestures and body language. All this is so that the machines can be accepted as companions for or complements to people.

Several factors are driving this acceptance in Japan. One, of course, is the rapidly worsening labor supply as their population ages and dwindles. Another is the Shinto belief that all things, even inanimate objects, have gods within them.

Many economies in the past were slave-based. In ancient Greece, it gave rise to an extraordinarily creative society. In other societies, results were less beneficial. In the 21st century, a new slave-based economy is developing, and we have to wonder what the consequences of that will be.

Slave-owning can have a corrupting effect, and in some ways it can rob both an economy and a society of energy and aspiration. Furthermore, machines, like humans, can be corrupted. Certainly, the persistent and exasperating problems of software remind us of that constantly. So does the common complaint from your banker -- "the computer is down."

The machines we're accustomed to -- dishwashers, washing machines, alarm clocks, automatic transmissions, etc. -- are clearly just inanimate objects. But as more of the machines we own are endowed with human qualities, whether it's appearance or ability to think, how will we see the relationship? How will they?

For managers of organizations, these are not just idle questions. The relationships in the work force between people and machines, already a somewhat difficult one, will become more complex and uncertain. Managers will have to think increasingly of how to get a specific task done rather than who will do it. Human Resource management will increasingly have to become Human-Machine Interface management. HR personnel will need new and different training and experience to do their jobs effectively.

Will emotion-equipped robots appear at some time in the future workplace in supervisory capacities? Should that happen, what near and long-term effects will we see? Will we be using robots to convey negative feedback, relieving managers of that task? And will this possibly mean people with poorer feedback skills can rise in management ranks?

Marketers will use methods that take full advantage of emotional interpretation by computers, conveying a sense of urgency in their marketing calls and messages, effectively pushing themselves to the top of a caller's priority list.

Emotive computing also raises the question of whether humans will become too dependent on computers to interpret emotions and feelings for them. As computers become increasingly ubiquitous and ambient in every aspect of our lives, our emotions will be evaluated and interpreted continually. For example, will we begin to rely on our answering machines or voice mail systems for determining the importance, or emotional context, of our messages? It is not unlikely, given the way in which people have come to depend on computers for correcting their spelling and grammar, and on calculators to verify their math. The other side of this concern is whether we want machines to interpret our emotional world for us, and make decisions about things such as when to interrupt us to take a phone call.

The fact is, people are not very good at seeing the consequences of disruptive technology. Who, for example, correctly forecast the impacts of the automobile on residential patterns, family structures and sexual practices? While many observers have attempted to predict the impacts of robots on jobs and work, only time will tell if any of them prove to be prescient. Meanwhile, all of us can be thinking about what the social and personal, as well as economic, outcomes might be.