Tuesday, November 20, 2012

Dickhead Roboticists

It appears that the fields of computer science and robotics attract a lot of Dickheads.  By that I mean fans of science fiction author Philip K. DickPsychology researcher David F. Dufty is acquainted with a few of these fans, academics, engineers, and artists, who undertook the strange challenge of building an android simulacrum of their hero.  He tells their story in How to Build an Android.


Dick is possibly the best choice for an author to imitate in robotic form.  He is well known for his story Do Androids Dream of Electric Sheep?, which was the basis of the movie Blade Runner.  The essence of the story is that robots have become so sophisticated, intelligent, and lifelike that it is almost impossible to distinguish them from real humans.  They blow away a Turing test, and the latest models can only be identified through a test that detects their carefully disguised lack of empathy; Dick thought empathy was an essentially human trait that may not be copied by computers.  The hero of Do Androids Dream must track down and destroy these nearly human machines.


Throughout this story and others, Dick explored identity and the shrinking difference between human and machine.  People suspect themselves to be androids, androids are programmed to believe they are human, and both have false memories that they cannot distinguish from reality.  A realistic android that interacted with people in conversation seemed like a natural fit with Dick.


Of course, the project had practical implications.  It required a lifelike interface.  Artist and robot maker David Hanson believes the face is a natural way for people to interact with each other and potentially with technology.  He has made a succession of realistic robotic heads that can imitate human expression.  The PKD robot was backed by artificial intelligence developed by researchers at the University of Memphis, particularly Andrew Olney, that allowed it to understand speech and construct responses based on things Dick had said and written in life.

To top it off, something happened in the brief life of the android that would have been well suited for one of Dick’s stories.  They lost Dick’s head.  Presumably it is still out there.  It may be in some airline warehouse.  It may be on the mantle of some thief, or some flummoxed buyer of a lot of abandoned baggage.  If someone has it, they haven’t been talking about it.

The big question of the book is how do we interact with technology?  What will it look like in the future?  We will be increasingly able to make machines that at least seem intelligent and look human.  Hanson thinks that is the way to go.  The Memphis scientists and engineers, and many other like them, continue to work on interfaces capable of increasingly sophisticated and seamless interactions with people.  Are we on a trajectory from Siri to Jacosta to C-3PO to R. Daneel Olivaw?  Are androids just what we need to make technology what we need or are they too creepy?

Dufty doesn’t pass judgment.  The PKD was an interesting piece of technology with and interesting history that raises a lot of questions about what machines can be.  Even as advance and capable as the PKD android was, it had a lot of limitations.  It could carry on a conversation, in a quiet room with a patient partner, but it did little else and its responses varied from the seemingly real thing to nonsense.

Personally, I don’t think we need, nor do I want, computers with faces.  I would be pleased if I could get my Windows 8 tiles to work.  I do think the way we interact with technology is important, especially that we are able to get feedback and other information from technology in ways that helps us understand what is going on and respond intelligently.

If you’re interested in this book, you may also be interested in

Dufty, David F.  How to Build an Android: The True Story of Philip K. Dick’s Robotic ResurrectionNew York: Henry Holt, 2012.

Google