Expand mobile version menu

Artificial Intelligence Specialist

Interviews

Insider Info

Some professions just have the ability to draw you in. By the time you're nabbed, you almost don't know what got you.

It was like that for Sue Abu-Hakima. Her parents wanted her to be a doctor. But unlike her sisters, she couldn't stand the thought of poking around in people's bodies. So she asked her brother for some suggestions. He was in engineering and suggested she do the same.

"He said it would only take three years, and that sounded good to me."

Abu-Hakima got her degree, but didn't stop there. She spent the next dozen or so years in university to eventually get her doctorate in artificial intelligence -- or AI, as it's called by those in the field.

"Once people get into AI they're hooked," she says. "And what we're seeing today is just the tip of the iceberg of what we can do in this field."

Abu-Hakima now works for the National Research Council as a researcher in the growing and profitable field of telecommunications. But she's also a group leader, mentoring other bright female engineering students and getting them interested in the complex world of AI. "I'm sure I'm going to be in this field until I'm a gray-haired little old lady," she laughs.

What's challenging about the field is the ongoing debate between industry and university. That's the discussion taking place between those living in the so-called real world and those who are wrapped up in the world of profit. That lack of understanding frustrates Abu-Hakima, who has straddled both worlds. "People are coming around to seeing industry's side, but it's taken a while," she says.

The other big challenge is getting a job at the end of the day. She's seen colleagues set aside all their paid work to focus on their studies. But when they finished their degrees, they weren't able to find work. "They find out that the skill set they came out with isn't as geared to the real world as they thought."

Will Neveitt revels in his work. He's an assistant researcher at MIT's artificial intelligence lab. Like others, he feels the area of artificial intelligence taps into some of the great mysteries of the world.

In his two years at MIT, Neveitt has worked on projects that would boggle most people's minds -- like teaching a computer to recognize people by the way they walk, to identify objects by their texture, and to build a virtual city out of information instead of bricks.

"You're continually challenging yourself to dream in bold, new ways," he says. "This is one of the great mysteries in the history of human imagination. It makes me excited to think that I'm working on problems that Plato and Aristotle struggled with!"

Fred Popowich is a past president of a national society for computational studies of intelligence. What he loves about this field is that you must marry diverse parts of knowledge together in order to achieve success.

It's not a simple question of screwing together a few nuts and bolts like some sort of modern-day Dr. Frankenstein. Instead, you have to be a keen observer of the human mind -- something Popowich enjoys immensely.

"But it's also a challenge. I'm always amazed at the vast amount of knowledge you have to get into a computer to do the simplest task."

To complete an accurate translation from one language to another, the programmer has to provide the computer not only with the complex ability to communicate in two different languages, but to also understand context. In some cases, that requires an understanding of the state of the world.

"There's no way you can get that into a computer. There are a lot of these things that we do so easily, but to get a computer to do them is tough. Data on Star Trek is a good example of that -- he's got the hardware, the memory requirements, the techno-babble and jargon, but we're not even there yet."

Popowich tried to get a computer to translate the lyrics to Elvis Presley's Love Me Tender. When the computer spat out the Spanish translation, it read "Love me, make me an offer" -- as in legal tender.

"Language is great," he says. "There's a lot of humor involved in this business. But maybe the way to judge a computer's success is if it can make the same mistakes as a human."

Tim Finin, a professor and researcher at the University of Maryland, believes language is the largest hurdle to jump when it comes to creating intelligent artificial life. That could take centuries.

Finin is currently working on a project funded by the U.S. National Institute of Science and Technology. He and his co-workers are trying to create software that will link up various computer systems to talk with one another in industry.

As an example, they're trying to link financial planning software with those used to order raw materials, schedule equipment upgrades and make purchase orders. If one part of production breaks down, the computers will be able to calculate how this will affect the bottom line and send out a message to purchasers.

"Right now, people are playing the part of the interface, so a lot of things may fall through the cracks or be inefficient," he says.

But Finin by no means believes computers are perfect. He recalls working on one program which would project global reaction to political decisions. That required a computer to have some common sense.

When the researchers posed a question to the computer and it didn't come up with a logical response, they had a look at what each of the thinking components of the software (called intelligent agents or softbots) was thinking about.

"It turned out that one agent was wondering if India might be just a kind of lawn furniture," laughs Finin.

"This stupid machine was using some plausible line of reason. The challenge then is how do you get a machine to have common sense? Or, for that matter, how do humans do it? It's going to take a long time to figure out."

There's no question that people in this field are always excited by the possibilities. But the frustration comes in realizing that, despite all the advances, humans still haven't been able to recreate the brain.

"It's a rewarding field because there's a sense that it's very cutting edge," says Finin. "The changes here are changing the world for people in a way that makes life better. It's a hot area and things are developing rapidly."

"It's a very creative field," says Abu-Hakima. "The only limit is your imagination. It's quite a phenomenal field. I've been in it for 12 years now and it's not letting up."