Robots are being designed to perform a broader array of work-related tasks. Global economic hardships may be (temporarily) causing the demand for industrial robots to decline,4 but improvements in artificial intelligence and the drive for efficiency will likely encourage companies to develop and use increasing amounts of robotic workers. Though the justification for automation is often couched in the language of liberation, this oversimplifies the complexities associated with technological change. Merely because technology is well designed from an engineering perspective, it does not follow that society's problems are solved. This is not to say that efforts to create robotic workers must stop, but the robotics community must be diligent in dealing with emerging ethical issues. Design pathways must be selected that either mitigate or prevent the negative consequences of using robots in the workplace. Otherwise, troubling historical occurrences, such as the decimation of certain segments of the work force, might be repeated.
With each significant technological change, visions of how improved and efficient our lives will become are typically offered. To some degree, the promise that we will be "liberated" from performing repetitive and mundane tasks has held true. Most of us do not mourn the passing of having to wash clothes or dishes by hand. Yet expectations in both our personal and professional lives tend to shift correspondingly, which in many ways counterbalances the "liberating" features that technology offers. Ruth Schwartz Cowan recognized years ago that the introduction of electronic devices into the home did not free women from the burden of doing household chores. As Cowan states, "What a strange paradox that in the face of so many labor-saving devices, little labor appears to have been saved!"1 In short, increasing expectations absorbed all of the extra time that was supposed to be freed up.
Similarly, we need to seriously consider how the increased use of robots will alter workplace expectations. For instance, if robots can help surgical procedures to be completed more rapidly, will demands on surgeons increase so they will have to perform more procedures per day? Expectations in terms of what it means to be a "good" professional are also likely to change, especially if a robot's error rate is lower than a human's. Briefly put, we should be wary of predictions that robots will be our liberators considering how the typical workweek does not seem to be getting shorter or less demanding in the digital age.
The U.S. military is enjoying the benefits of robots since they can complete "dull, dirty, or dangerous" tasks, and their labor is very useful in the civilian realm as well. Yet automation can eliminate job opportunities and usually causes the demographics of the work force to be significantly altered in a relatively short amount of time. Employers find robots to be rather enticing since they do not receive benefits or request vacation time. Through the design choices they make, scientists and engineers play a key role in determining the kinds of employment practices that can and will transpire.
If categories of jobs do indeed vanish as a result of robots, will the relevant skills of displaced workers be transferred to another application or will those skills be rendered obsolete? This concern is not unique to robots. But what may be a new variation now is that the jobs available to humans may be drastically reduced as computers, the Internet, and robots replace humans in employment sectors that used to be thought of as immune to automation. At present, it is fairly difficult for people to find work that is not connected in some way to these technologies. This development might not be conducive to the flourishing of each person's respective talents, and robots are likely to exacerbate this situation. Also, the type of skills that will be in demand if and when the robotic age takes hold might be obvious in some ways but not so apparent in others.a
The impact of robotic workers can and will extend beyond the elimination of labor-intensive jobs, which captures a key reason why the ethical dimensions of robots seem to be drawing increased attention. It is not only possible to eliminate "dangerous" and "boring" work but at least some jobs requiring specialized expertise, such as being a surgeon, may start to disappear. A decade ago, Bill Joy, the co-founder of Sun Microsystems, famously warned against this.2 Even if we don't share Joy's apprehension about the future of robotics, we can still appreciate the perils of trying to replace "uniquely human" abilities such as critical thinking and intuition.
Scientists and engineers should reflect on their ethical responsibilities to communicate with the public about a robot's capabilities and limitations.
To illustrate this point, we can look at the robots being created to assist with the health care needs of elderly populations. An outgrowth of this effort is that it could subtly or perhaps dramatically change how nursing homes function. In principle, robots could free up the time of nursing home staff; for example, a robotic assistant can provide medication reminders or warnings if a resident is in danger. Such a robotic counterpart might enable human workers to be more caring and productive. However, nursing homes and other care facilities will be tempted to downsize their human staff when a robot is "hired" instead of freeing up human staff to give more time to residents.3 Since many nursing home residents in the U.S. and elsewhere already do not get enough care and individualized attention, this is a very troubling possibility. Theoretically, an increased emphasis on in-home care could for example lead to the creation of other types of jobs but we should be skeptical about this. Financial considerations, the drive for efficiency, and overconfidence in technology are strong driving forces that can push humans "out of the loop."
On a related note, reliance on automation may exacerbate a common human tendency to shift our attention to a different task when we believe (perhaps falsely) that we can trust someone or something else to deal with the task at hand.b Returning to the issue of health care, will nursing home staff be less attentive if a robotic assistant is placed in a resident's room? The more reliable we think automated systems are, the more likely it is our attention will stray. What complicates matters is that this type of behavioral shift might not be consciously detected. Hence, it would be wise to temper the confidence that users place in robots and other automated systems, especially when people could be significantly harmed. This could be accomplished in part by ensuring that risks are transparently presented to users. To that end, scientists and engineers should reflect on their ethical responsibilities to communicate with the public about a robot's capabilities and limitations, and not merely leave it to marketers, sales departments, and others to fill this role.
Ethical concerns about integrating robots into the workplace are becoming increasingly pronounced. Again, the intention here is not to stop innovation. Rather, the hope is to inform the design process. Ideally, the robotics community will select design pathways that mitigate the associated concerns and thereby enhance the public's lives.
4. Tabuchi, H. In Japan, machines for work and play are idle. The New York Times (July 12, 2009); http://www.nytimes.com/2009/07/13/technology/13robot.html
a. For example, in Wired for War, P.W. Singer discusses how cooks might have more job security than military pilots because they can prepare food in creative ways. In the civilian realm, he reassures hairstylists by suggesting their specific abilities may keep them employed; Penguin Press, NY, 2009, 130132.
b. Placing too much confidence in technology, often at the expense of other sources of information, seems to be a growing problem with GPS in automobiles; see for example, Is your GPS navigator a friend or foe?" The Sydney Morning Herald, (Jan. 12, 2010); http://www.smh.com.au/executive-style/gadgets/is-your-gps-navigator-a-friend-or-foe-20100112-m4ei.html
The author would like to thank Rachelle Hollander, Keith W. Miller, and the anonymous reviewers for their helpful insights and guidance.
The Digital Library is published by the Association for Computing Machinery. Copyright © 2010 ACM, Inc.
No entries found