In science fiction, human beings are frequently pitted against a metallic and mechanized other—manifestations of our culture’s fascination with and fear of technology—from the ominous robot Gort in 1951’s The Day the Earth Stood Still to the various iterations of the Terminator. Robocop, both the 1987 Paul Verhoeven version and the recently released reboot, locate that conflict within the body and mind of a single individual. Is Robocop, as the new film asks, a man who think he’s a machine or a machine that thinks it’s a man?
Science and Film asked Ayanna M. Howard, Motorola Foundation Professor and Associate Director of Research at Georgia Tech’s Institute of Robotics & Intelligent Machines, to parse out the scientific plausibility of “Robocop” and to help us distinguish the dividing lines between humans and robots. Howard has worked with NASA on various robotics projects and received worldwide attention for her SnoMote robots, designed to study the impact of global warming in Antarctica.
Sloan Science and Film: Before we get to Robocop, what are you currently working on?
Prof. Ayanna Howard: When I came to Georgia Tech in 2005, I went into healthcare and focused on the question: how do you make robots interact with people? Now you have robot clinicians, robots as exoskeleton devices and robots helping individuals with disabilities, working closely with them in the same frame of mind, though they can’t take full control. Because healthcare is very sensitive, you want the robots to be efficient and effective, but you don’t want them violating rules or regulations. On Mars, for instance, if a robot does something bad, it can cost a lot of money, but if a robot does something bad to an individual, that’s a whole other issue.
SSF: What could a robot do that is “bad” involving a person in a healthcare role?
AH: Applying too much force, for example. If I’m interacting with a child and doing rehabilitation, I might be gently holding her arm and helping with range of motion. But if you do that too forcefully, you might break something.
SSF: So if you extrapolate from the science we have now, could you foresee creating a cyborg based on the human brain as seen in Robocop? Could you attach computers and robotics and make something move and think and police?
AH: I think so. Not in the next five years, not in the next ten years, but in my lifetime, I think it’s possible. If you look at the science we have now, we have artificial legs, artificial arms—we can have two arms up to the shoulder and two legs up to the hip being replaced. We now have artificial hearts; we have research on an artificial liver. So the only thing that’s missing is the spinal cord. But there has been research tapping into the brain. You now have paraplegics operating robot arms connected directly to their brains. So I think the technology is there. The issue is medical: Can you keep the brain alive? Can the brain still function? How do you keep the blood going to the brain? But I think you could control fully autonomous robotic appendages with the human brain.
SSF: How do computer scientists meld computer chips with brain synapses?
AH: It’s invasive. They’re cutting into a person’s skull and putting in probes. Some of the work where they are adding robot arms to an appendage, they’re looking at the nerve and the signals of the nerve. They have to grow the nerve, so there’s live wire that you can connect to the robotic platform.
SSF: In Robocop, there is a scene in which a classical guitarist is playing with two robotic hands. I wonder if such a robotic hand could be finely calibrated enough to play classical guitar?
AH: We don’t have one right now. There has to be a feeling in the nerves that provides feedback, you have to feel the vibrations of the string, which would push you to go to the next note or hold it a little tighter. That’s the sensing. Then there’s the dexterity--allowing the fingers to move at a certain speed and a certain range. Both of those are important. We are getting to the point where we have artificial skin, where you can feel things over larger areas; I can brush up against something and feel if it’s smooth or if it’s rough. In terms of extremities, there’s research into micro- or even nanosensors, which means we can now create actuation that is much smaller, so it’s not as clunky and it’s faster. But the real question is how do you control those degrees of freedom and how do you connect the human to those degrees? And that’s where the real hard science is going to be.
SSF: In the film, and in a lot of sci-fi, there’s this theme that comes up about the consciousness of robots. As a robotic scientist, do you ever consider these ideas? Where is the line between consciousness in humans and in robots?
AH: I am a robotics person, but I am also a devout Christian, so I think there is a limit to how far we should push robots. Can a robot be a sentient person? Unless we introduce human parts, I don’t think we can get to that level. Having a brain attached to an entire robotic platform, I think that’s still a human, whereas vice versa, having a robot brain and attaching human parts, I don’t know if you consider that a human. If we push robotics to the point where we are unable to distinguish whether it’s robot or human, there are so many ethical issues that we are not able to address right now.
SSF: There’s a point in the film where they’re trying to decide: Is Robocop is a machine who thinks it’s a man or a man who is a machine? What’s the difference, in your opinion?
AH: I think what makes a person is in the brain. I think the growth of the brain and life experiences give us consciousness and awareness of self. If you think of the difference, Robocop is human, even if you try to “erase’ his mind. I think there’s something inherent about growing up, and having the brain evolve in a human environment, which I don’t think you can take away from someone. If you look at a paraplegic—they have no functions, they’re not touching, they’re not feeling, they’re not walking, but no one would argue that they’re not a person.
SSF: Is there anything that needs to happen, technologically or scientifically, to get us to the next level of robots?
AH: I think there’s lots of different progressions that are going on right now—miniaturization, that’s a requirement. Things such as getting more intelligence on a chip, more processors on a smaller packet size, making actuators that are miniaturized, but also powerful. The problem with a lot of our systems now, for example, is the way you create a finger now, you have the motor outside of the joint, so you might have a cable in the way. But if you have the motor in each of the joints, you get a more equivalent range of motion and ability, so it shifts the way we design robotic systems to simulate the actual movements of a finger.
SSF: When you’re called to Capitol Hill in the future to testify for or against Robocop, what would you say?
AH: I would say Robocop is a good thing. Because when we design good technology, there are so many positive benefits. There might be one rogue Robocop, but all the technology that allows you to create Robocop also allows a child who has been in accident to still have a life, or someone who has had a stroke to still have a life. I think it’s a positive thing, and I think that research still needs to continue.