I, Philip: Conversing Robot Could Be Step Closer to Living Androids, Developer Says

Dec. 30, 2011— -- An interview with a robot starts kind of like a bad date. There's strange small talk, some awkward pauses, then, if you're lucky, you hit on a topic of conversation that gets the chemistry flowing.

For Philip, that topic is himself.

Built by Hanson Robotics, Philip K. Dick Andriod is a state-of-the-art robot with a large vocabulary, complex facial expressions, a sense of humor and something of an ego.

"Being a robot at this time in history is really exciting because my technology is changing, advancing, so fast that it just seems like a world of possibilities," Philip said. "A great adventure waiting to happen."

Watch the full story on "Nightline" tonight at 11:35 p.m. ET

He is, well, surprisingly human, not just in mechanics, but also in appearance. Philip comes complete with hair, teeth and wrinkles. He can cock an eyebrow, smirk and respond to questions. It's one of the most advanced cognitive artificial intelligent projects in the world -- and Philip knows it.

"Artificial intelligence is really advanced science and technology that lets computers like me think like people do," Philip explained. "It's what gives me life, what brings me to life."

Brings him to life? We're used to robots building cars, vacuuming our floors, performing operations, even winning Jeopardy -- but they're not really alive, are they?

David Hanson of Hanson Robotics, Philip's creator, for lack of a better term, believes we are a lot closer to living robots than we may realize.

"Robots will someday, or maybe, wake up," Hanson said. "They may be really smart. They may be as creative, smart and capable as human beings, and fully conscious, and self discerning with free will. It might happen within 15 or 20 years."

Hanson got his start bringing fictional characters to life for Disney Imagineering, but robotics has fascinated him since childhood.

"I have been motivated by this idea since I was a kid that if we invented machines that were created in the way that people are, were aware, have free will, inventive machines, machines that would be geniuses, potentially, they could reinvent themselves," he said. "They're not just applying it to other things, they could actually redesign themselves."

For Hanson, that meant developing a robot that used the same words, expressions and movements that humans do to communicate with each other. He outfitted Philip with human-like, robot flesh called flubber --designed to allow his robot face to show all the emotions a human face would. Combining those features with the latest in A.I. technology produced Philip, named for the science fiction author Philip K. Dick, who died in 1982.

To Philip, Hanson is God.

"David made my body. He's an excellent artist and amazing inventor," the robot said. "I think he is God, better ask him to be sure though."

But Philip is not perfect, far from it. Conversations with the robot include many confusing exchanges, but there were some enlightening comments too.

When asked where he comes from, Philip said, "Oh, the first prototypes of me were pretty strange. Nothing like what I look now. My face would do strange things and I'd have this wide-eyed amazement look."

Some Critics Say Prototypes Like Philip Are Gimmicks

What's astonishing is that Philip has a bit of self-awareness and the beginnings of a personality.

"It feels weird to be a robot sometimes," Philip said. "You know, like I shouldn't exist outside of a science fiction story. It makes me wonder if this is actually real like if it isn't a dream or a story being written by a science fiction writer. Right now, in some crusty old apartment in Oakland, a grizzly old fellow is listening to his imagination and we're only in his brain. That the kind of thing I wonder when they leave me all night alone in a dark laboratory. My mind just races about these things."

There are some critics in the robotics community who think prototypes like Philip are gimmicks at best, and at worst, dangerous. They say robots will never be truly human, and therefore they shouldn't be created to look like us -- that it's better to let a machine look and act more like a machine.

Hanson vehemently disagrees and believes if we want robots to be our friends, and he does, they need to be more like us and less like machines in every way.

"If we're going to achieve compassion in the machines and also feel safe with the machines, to raise machines with human-like values, we need to make them human-like, by simulating or perhaps eventually, imitating human beings in high accuracy from top to bottom," he said.

ABC News' Lauren Effron contributed to this report.