The consequences of technological change are never fully predictable. ICT-based systems yield unanticipated consequences, including "revenge effects." They can also suffer from "brittleness," breaking down when faced with situations not anticipated by designers. The following sections address the long-term, large-scale impacts of new technologies on organizations and on society as a whole.
ICTs designed for one purpose can lead to a number of unanticipated "second-level effects,"24 often resulting from individual and organizational reactions to the new technology. Some unanticipated effects are pleasant surprises. One example is Internet email. Designers of the early Internet never imagined that email would become the Internet's most important use. Users, reacting to the opportunities created by the early Internet, wrote the first email programs and transformed the system into a new communications infrastructure.25
However, other unforeseen effects are shocking disappointments, or worse. The 1987 stock market crash — among the worst in history — resulted in part from the introduction of automatic, computerized trading systems by large investment firms. These systems responded to a sharp market downswing by rapidly and automatically selling off large volumes of stock, flooding the market and driving prices deeply downward, turning what would probably have been no more than a bad day into a catastrophe. "Revenge effects"26 occur when unanticipated consequences produce an effect diametrically opposite to designers' intentions. As one example, in the late 1970s computer manufacturers trumpeted a future "paperless office." Yet the computerized office of today consumes considerably more paper than did its 1960s counterpart.27 Computers make it much easier to copy, revise and distribute copies of documents, and individuals still often prefer working with printed out versions of those documents. Many existing practices, procedures and policies still tend to rely on hard copies. Several recent studies would also seem to suggest that paper is still superior to existing computer interfaces for supporting many common office tasks.28
Systems relying on ICTs pose some unique risks. They can serve as excellent tools to supplement human skills. But when we attempt to automate sophisticated human activities, the results can often be disappointing. ICTs do not truly understand the world as we do. As a result, they may prove too brittle to accommodate novel situations.29 Another risk involves "deskilling." Replacing human presence with remote control can mean that people lose a direct, intimate awareness of events as they occur. Airline pilots and factory supervisors, for example, must now work through numerous layers of technology, separating them both physically and experientially from the equipment for which they are responsible. This can sometimes make it more difficult for them to identify and appropriately respond to problems that may arise.30
Predicting and controlling the operations of complex systems is extremely difficult, and all the more so when such systems are constantly changing. A large literature on "complex adaptive systems"31 investigates ways to think about such situations. Designers can often reduce risks and unanticipated consequences by not venturing too far from proven designs which have worked well in the past.32 Yet the power of ICTs creates a strong temptation to build completely new methods, ignoring the lessons of the past in the search for a revolutionary future.
» Next: An Example: The Productivity Paradox