“Hitachi’s new toddler-like robot rolled around and waved for reporters on Wednesday [November 21], only to crash into a desk and demonstrate the challenge of turning automatons into everyday helpers. The red and white robot, designed to run errands in offices, wasn’t prepared for the jam of lunch-break wireless network traffic at the company’s research center. Unable to communicate with its handler’s laptop, it smashed into the office furniture as reporters gasped. Reporters had to wait for an hour until after the lunch break to watch the robot repeat the demonstration — this time smoothly making its way between the desks,” according to an Associated Press report in the Washington Post by Yur Kageyama, November 21, 2007.
In September, I visited the open house at the Intel Research Labs in Seattle. The scientists were displaying their latest research at various stations around the room. Lots of exciting ideas were being demonstrated, but there seemed to be common themes. Position sensing, wireless, and reasoning capability were all built into demonstrations to locate and communicate with people and things.
At the MIT Media Labs (see the update on Roz Picard’s work on p. 17) we see the emphasis on “Things That Think,” drawing on advanced algorithms and the increasing capability in computing to do things that could not be done before. The work in affective computing (related to computer emotional recognition and response) is now coming into some exciting areas of potential practical use.
I have four reactions to this wave of new technology. The first two relate to the technology itself, and the second two relate to its business application.
What Is Different Now?
1. Haven’t we been here before? Many similar things were promised in the last great wave of artificial intelligence (AI) in the 1980s. The science-fiction dream of having computers “come alive” was full of promise then. But after the technology community promised so many things and failed to deliver, even the term AI fell on hard times.
There are two things that are different today. The first comes from Moore’s Law, which we discussed with Pat Gelsinger, senior vice president of Intel Corporation, in the Conversation in this issue. Because of Moore’s Law, we have an underlying computing capability that is 100,000 times greater than it was 25 years ago. These problems were much harder than they looked in the 1980s, but the more powerful capability from computing provides a set of more powerful tools to address them.
The second comes from maturity in the computer science field. Computer science was a teenager in the 1980s, making immature promises it had no basis for making. The field has grown up. Many of its leaders are more careful now, because they have a better understanding of just how difficult some of these problems are. You can see some of this in the discussion with Rosalind Picard on p. 17.
Will the Technology Succeed?
2. Will all of this work succeed? Of course not. If all research succeeded, it would not be research, but product development. The problems in AI remain deceptively easy to describe and deceptively difficult to solve. Recognize speech. Perceive emotion. Do reasoning. Further, AI problems have always been plagued by this: Two seemingly similar problem statements represent a very easy problem and a very difficult one.
Some applications for robots will succeed, and some will fail. Some research combining positioning and pattern recognition will address real problems when it grows up, and some will only address “toy” problems. Emotion recognition by computers will find a place somewhere in business, but where is not yet clear.
But I think there is growing evidence that more and more of this work will succeed and find its way into real applications.
The Business Adoption Challenge
3. These new technologies raise a new wave of challenges for business. It has always been true that some technology offers truly revolutionary transformation, and other technology offers only higher costs and frustration. It is often a long path from the results of research to the results for business success.
We saw this in the last decade as businesses sorted out the role of the Internet. Some saw the Internet as the solution regardless of the problem, and we got things like pets.com. Some saw it as a new way of doing business and we got amazon.com or Google. The technology made it possible, but the wise application of the technology transformed the business.
But I believe the application of these AI related tools will be much tougher for businesses to adopt than the technology of the past. PCs changed the life of administrators. They also changed the life of an engineer. The company leaders could look for efficiencies and cost savings from this work. Enterprise systems changed job descriptions and life throughout the “back office,” and even changed the customer-facing tasks. But for the leaders, more information was available from which to make decisions. The supply-chain management software enabled outsourcing of manufacturing. Many manufacturing jobs went away from the West. The leader’s job, and pay scale, often grew as the virtual empire spread throughout the world.
These new tools could be different. Might they mean that more strategic thinking can be done by the computing system, making some of the leaders less important? Might they mean that the executive coach could be replaced by a computer, making coaches redundant? How will the “technology resistance” be different with this wave of tools than it was with computer aided design, PCs, enterprise systems, and supply-chain management? I think they will infringe a bit more on the province of company leadership and their role is not as clearly defined for business. Together this will mean heightened resistance and an increased likelihood that they can be dismissed.
When I was involved in managing a technology R&D organization, we identified some real irony. Through the technology we developed for the company, we were changing the way of doing work throughout the company, and wondered why there was so much resistance to our great ideas. But when our world was changed, we showed the same resistance ourselves. I am wondering how these tools will affect the leadership and what this will mean for adoption rates.
New Ethical Concerns
4. New technology always takes us to new territory where we must rethink what is right and wrong, what is ethical. Connectivity through telecommunications opened the door to remote access. But it opened the door to hackers as well. Cell phones introduced mobility for workers, and closed the door on private and personal time in the way it used to be. Digital technology introduced products with almost no manufacturing costs, and introduced a new kind of piracy for software, music, and other products that are made up of “bits.” Computer graphics made visual images come to life for games and for computer-aided design, and made possible violent games and pornography.
Our tendency is to think about what technology will do for us, and pay less attention to what it might do to us.
There are some obvious things that might be said about the newer technologies. A robotic office assistant that does the errands around the office could be a big help. But as we saw earlier, products are not completely reliable. The problem is, they never will be. Something about the network caused the robot to “lose control” and run into a desk — could it also run into a person? How do office dynamics change when the robot replaces a few people? Does the office robot get invited to the office party? Do we ever get up from our computers if the robot can do all of the running around for us? And if we don’t get up from our computers how will this affect our thinking ability, not to mention our waistlines?
Products related to mobile workers suggest all types of privacy and hacking possibilities. And they also raise again the question of personal time vs. company time.
How might the computer reading my emotions be abused? Can I go to jail for what I am thinking, or what my face indicates I am thinking? The science fiction movie Minority Report raised this question when the police arrested someone “for the crime they were about to commit.” It was only science fiction, but will we be getting closer to this in the office of the future? What happens when the computer gets the emotion wrong? This statement came from one of my books of quotes: “If I look confused it is because I am thinking,” Samuel Goldwyn. If people get it wrong, what about computers?
Conclusion
None of these considerations dampen my enthusiasm for the technologies that are now in the laboratories. They represent exciting technical work showing promise to make our lives and our businesses better.
But if we have learned anything on this technology journey, we should have learned that for every upside of technology, there is a downside. Further, complexity often finds a way to cancel even great progress. So this suggests we should start early anticipating and mitigating the potential downsides of this next wave of technology. As long as Moore’s Law is alive, this wave will continue. But the underlying continuity masks the changing nature of the products that will affect business, and it will take new skills to deal with both the technology and its impact on people.
Al Erisman is executive editor of Ethix, which he co-founded in 1998.
He spent 32 years at The Boeing Company, the last 11 as director of technology.
He was selected as a senior technical fellow of The Boeing Company in 1990,
and received his Ph.D. in applied mathematics from Iowa State University.