Evolutionary Learning/Rights for Robots

Wednesday, crazy busy day. Work projects, getting ready for my trip, a Landmark conference call, Kaveh at noon (our last phone chat before I go down there on November 5th for the big good-bye which has been in the works for a while), and more. Good to be busy! And writing group tonight which will be Liza's last (from now on she'll be in Urbana during the week). Should be interesting too in that James and I haven't spoken since his rude text (not sure how it's going to play out).

Last night Catherine and I at Hamlet - Writer's Theater in Glencoe. She loved it - I didn't. I found myself dozing off, escaping the long confusing speeches. I love Shakespeare - used to have season tickets at the Chicago Shakespeare Theater, so not sure why I had such a hard time with this production. Either my Shakespeare language synapses in my head have atrophied or I was just too tired to handle the Middle English. And Hamlet - I found him annoying, slight and effeminate, whiny. He pranced around the stage, scowling most of the time, flinging scarves over his shoulder. Not a manly prince at all!

Feeling like a kid in a candy store this morning - so many fascinating articles to read in New Scientist and not enough time. Seriously! What's not compelling about titles like, Bionic Legs Could Soon Be Controlled by Thoughts, Handle the Truth: Navigation Knowledge in the Web Age, Can Castration Really Prolong a Man's Life?, New Killer Virus in the Middle East, Our Token Involvement in the Game of Life, and finally the one I just had to read, Mimicry Beats Conscious Gaming's Turing Test.

Last we checked in with the Turing test, we read the book called, The Most Human Human by Brian Christian. The Turing tests are events where robots and humans vie to be dubbed, "The Most Human Human". Based on scientist Alan Turing's work in the 1950's in which he asked the question, "Can Machines Think?", the test measures a machine's ability to exhibit human behavior. Until now humans have always won the Turing tests, barely scraping by in recent years as robots have become more and more sophisticated. Now, for the first time, the bots have won - in a recent test where bots played video games alongside humans, the judges incorrectly chose the bots as the human players.

And so interesting to understand what some of the big challenges the bots had to overcome! Leaving a room for instance - you don't even think about it, right? For a bot, maneuvering out of a room is a very tricky affair, and until now, the quality of their movements towards the door have been a dead giveaway to their machine-ness. To overcome this, the winning bot was programmed to study and mimic their human opponents behaviors, even though much of what we do must seem illogical to a robot.

His team recorded dozens of humans playing Unreal Tournament, spliced out the bits where the human released themselves from different geometries and then programmed the bot to deploy the human strategy used in a situation most similar to whatever it finds itself in. For other aspects of play, UT^2 deployed evolutionary learning, in which successful strategies were bred to produce offspring strategies that were even more human-like.

Article concluded that, now that the spatial reasoning barrier has been cracked, the next big breakthrough will have to be language.  

Still, Miikkulainen says that intelligence is made of up of components, at least one of which has now been solved. "In terms of spatial reasoning, it is possible to act human," he says. "Language is a much bigger problem, but it's nice to know that in this part of intelligence, we can do well."

The film critic Roger Ebert would put robots to a totally different kind of test. No robot has yet been able to pass the Ebert test which would measure a robot's ability to tickle our funny bones. His test would determine whether "a computer-based synthesized voice has sufficient skill in terms of intonations, inflections, timing and so forth, to make people laugh." (Wikipedia)

This topic is fun to read and think about. Fascinating, right? These tests - academic fun but not really affecting our lives in any appreciable way. But I think that's going to change in the not too distant future. We seem to be on a trajectory where the barriers between humans and machines are going to continue to be challenged and breached. I predict this topic will become very heated with people facing off - the naturals versus the transhumanists (cyborg-wannabees). 

Challenge today is thinking about this. We saw last week that the first Transhumanist politician has been "appointed" There are ethics boards springing up, attempting to apply human rules and values to the emerging science. There are even people concerned with robot rights - seriously! Some folks feel like robots could become an exploited underclass and as such their rights need to be protected (robots sharing our "inalienable right to the per-foot of happiness?)

Someday robots will be in our houses as playmates for children, servants for adults.  They may become sex surrogates.  They will be in the courts as judges. They will be in hospitals as caretakers.  They will proform dangerous military and space tasks for us.  They will clean pollution, save us from numerous hazards.  The child who loses her robot because of malfunction will when she grown up always remember her robot.  She may, at the insistence of her parents, relegate robots as persons of the world of fairies, goblins and ghosts, the unreal and the impossible. Or she may decide that her robot like her family, friends and pets is part of her, is part of life itself. 

We believe that robots will one day have rights.  This will undoubtedly be a historically significant event.  Such an extension of rights obviously presupposes a future that will be fundamentally different from the present.  The expansion of rights to robots may promote a new appreciation of the interrelated rights and responsibilities of humans, machines and nature. (The Rights of Robots: Technology, Culture and Law in the 21st Century by Phil McNally and Sohail Inayatullah.)

I'm sure too there are corporations whose fortunes are being hitched to the star of transhumanism.  Robot soldiers, robot teachers, robot doctors - far fetched, right?  Nope, they're already here. Are we OK with that?

Peace,
Sarah