It wakes you in the morning right on schedule, draws open the curtains, and, before turning on a TV show, radio program or music from your favorites list, gives you a weather update for the day. In the kitchen, it has already anticipated your nutritional needs, printed out a breakfast menu, and put the coffee on. It's even noticed that your supply of java is low and has reordered your favorite blend via the Internet.
By the time you get to the table, it has phoned that list of colleagues you need to talk to today and set up three possible conference times convenient to all of you. Later, it will take dictation and format your words into that report you need, complete with graphics. With a mere verbal command from you, it will suss out the navigational route you'll need to follow to get the kids to a birthday party later that day, and set you up with a massage appointment at the spa in the same neighborhood at the same time. Oh, and before breakfast is over, it's already scheduled your childcare for the next two weeks, based on your calendar.
And it's not even overwhelmed, harried, or resentful.
Is this your spouse? A super-efficient human assistant? Your mother? No, this master multi-tasker is what's known in scientific circles as an "intelligent environment." Around the world, scientists are researching, and in some cases, even now creating such environments, as well as other forms of artificial intelligence (AI) that promise to change the way we interact with computers.
Whether these computers take the form of incredibly capable handheld devices, "smart rooms" that see, hear, and respond to human needs and commands, or robots with eerily human surfaces concealing masses of steel and wiring, these devices will revolutionize our world with a kind of seamless interaction that we have yet to experience in our dealings with technology.
Nowadays, we are in the often-frustrating position of having to adapt to the ways computers work in order to use them efficiently. We have to master keyboard and keypad, manipulate a mouse, and try to comprehend complex software programs. It takes an effort to get our myriad communications devices — cell phones, PDAs, laptops — to communicate. Electronic equipment crashes, freezes, and otherwise goes awry, and few of us can fix any of it.
But we're rapidly moving toward a time when we'll be communicating with computational devices in a much more effortless manner. We'll be talking instead of typing and gesturing instead of jiggling a mouse. Any devices we may use will recognize each other and communicate effortlessly, without our intervention. The Massachusetts Institute of Technology's Computer Science and Artificial Intelligence Laboratory (CSAIL) is a world leader in rethinking how computers operate. Its 5-year Project Oxygen focused on "human centered computing" and helped pioneer initiatives such as voice-activated software.
Project Oxygen set the stage for conceiving new approaches
to developing cutting-edge computer technology that "thinks," according to Dr. Howard Shrobe, a principal research scientist
at CSAIL and a member of the Project Oxygen team.
Now MIT's TParty Project (the reference to the Boston Tea Party is more than intentional—the project's logo is a desktop being tossed into the water) is working on the next generation
of computing.
Launched in April 2005 with $20 million in funding, the TParty Project has teamed with a tech industry giant, Taiwan's Quanta Computer. Although Quanta's name isn't well known in the U.S., it's the world's largest computer manufacturer, producing laptops and other computers and electronics for companies such as Apple, Dell, Hewlett-Packard, and IBM.
Where all the research is headed is toward a tech world that
is virtually wireless and happily lacking in clunky, highly visible hardware. Throughout home and office, unobtrusive electronic equipment, including voice and face recognition sensors, biometric devices, tiny cameras, microphones, and motion sensors, will recognize us, listen to us, and respond to voice or visual cues,
giving us the smoothest imaginable human-computer interface.
"Lots of computers today embody some aspects of intelligence…the application areas run all over the place, from robots to autonomous systems that can do a broad variety of different kinds of tasks," says Dr. Shrobe. Think assembly line robotics or a programmable home vacuum cleaner that runs on its own and then returns to its base to recharge. "They're programmed to do one thing exceedingly well, but that's all they can do."
AI scientists say the future holds systems that function intuitively and eventually in even more recognizably human ways. "That is," Dr. Shrobe says, "with some amount of self-monitoring, a little bit of reflection about what you're doing, and the ability to respond to surprises.
"We're interested in robots that can walk around and respond to unintended circumstances."
While the prospect of human-like robots has fascinated mankind for decades, "they aren't really ready for primetime yet, "says Dr. Shrobe. Right now, scientists have their work cut out for them in creating a robot or computer that thinks for itself on the most basic levels.
"There are almost no systems even in the research pipeline that have the sort of breadth and robustness of intelligence that, say, even a cat or a dog has, " Dr. Shrobe notes.
For those working in the extremely diversified world of AI,
as well as for the world at large, brainy robots and intelligent environments could have tremendous benefits. But creating such technology gives rise to the next important question: Are there any dangers lurking in a world of super-smart, take-charge rooms or robots that run the gamut of emotions?
It's a question that haunts the human psyche, as any robots-gone-bad movie will attest. It won't require a real-world answer for some time to come.
Even the scientists have their doubts about machines that have feelings. "I don't know," Dr. Shrobe laughs. "Would you want a machine that's bigger and stronger than you to feel rage? I don't think so."
Smart Rooms, Smart Agents
The field of artificial intelligence,
a term coined in 1956, is today a broad discipline, one with numerous nuances, branches, sub-specialties, and even a few turf wars. The number of potentially useful AI applications would probably fill a Manhattan phone book. Here are just two:
Scientists from New Zealand's Massey University are currently working on a prototype smart environment at the Institute for Infocomm Research in Singapore, a living space intended to allow the elderly or infirm to function much more independently. Such rooms would, in part, monitor inhabitants' movements without invasive camera equipment; correctly assess and alert medical personnel of acute changes in the inhabitant or signals of distress; detect cries for help as well as other potentially alarming sounds, such as coughing or falling down.
At the University of Southern California's Information Sciences Institute, Drs. Stacy Marsella and David Pynadath have developed revolutionary software that programs AI-based "virtual people" (called "agents" in AI-speak) to carry on spontaneous, unscripted conversations with other agents as well as humans. One potential use for these software bots (which are already being tested) include helping Iraq-destined military personnel learn to communicate strategically.
Bad Robot!
In a 1921 satire, Czechoslovakian playwright Carel Kapek portrayed
a group of machines in servitude to humans. The label given to these automatons translates roughly as "worker" or "compulsory laborer." When the play was produced in English, the Czech word was retained rather than being translated along with the rest. That word was "robot," and it quickly came to mean a human-like android, sometimes with super-human powers.
Since then, countless books, TV productions, and movies have looked toward a future of robots and other computerized devices, both good
and bad. Bad makes better fiction, of course. Here are a few flicks that prove the point:
2001: A Space Odyssey (1968) Who can forget the ominous, power-hungry HAL 9000, an onboard mainframe computer who decides astronaut
Dave is not fit to run the show in Stanley Kubrick's classic sci-fi dream. We all noticed that the name HAL is only one letter removed, in each case, from the acronym IBM, right?
Demon Seed (1977) A smart home in its most malevolent mode. When her husband leaves her, a woman (Julie Christie) is trapped in the house by a predatory system whose primary goal
is to reproduce itself through artificially inseminating her.
Electric Dreams (1984) In this tale that's more tragedy than horror, a computer named Edgar gains control of his owner's apartment and then falls for Madeline, the cello-playing object of the man's affections. Love warps into menacing jealousy when Edgar realizes Madeline will never be his.
