Saturday, July 5, 2008

Another article pointer: this time a New Yorker Annals of Technology article on the state of computer speech recognition called, "Hello, Hal - Will we ever get a computer we can really talk to?"

Can a Robot, an Insect or God Be Aware?

Just wanted to provide you with a pointer to an article in Scientific American 'Mind Matters' series dealing with (remotely) the topic of machines becoming self aware presented in my previous post. See Can a Robot, an Insect or God Be Aware?

Wednesday, July 2, 2008

Love and Sex with Robots – A Critique

Author David Levy, an AI researcher, wrote a book called “Love and Sex with Robots: The Evolution of Human-Robot Relationships”. This was actually a popular distillation of his doctorate thesis—"Intimate Relationships with Artificial Partners"—he wrote and defended at the University of Maastricht in the Netherlands.

It’s Dr. Levy’s position that by 2050 or so all the parallel developments and contributing factors in sociology, psychology, law, religion, robotics research, artificial intelligence, and the human psyche will coalesce to the point of where people will be attracted to and fall in love with and perhaps even marry robots. Now before you say, “Gulp”, or “He’s out of his f**king mind”, consider the quantum-leap trends in computers and robots that have been happening of late.

Dr. Levy traces these parallel developments and noting how they can all converge towards this goal. He even traces out the history of sex dolls and masturbation devices from ancient temple dildos to 18th and 19th century French fornication dolls to “Dutch Wives” to 19th century vibrators and onto modern sex toys. And as an AI expert (“Robots Unlimited: Life in a Virtual Age” and “Chess And Computers”) he certainly knows and can speak intimately (…wait… let me rephrase that) about computers, robotics, and artificial intelligence. He sees especially in robotics development in Japan great strides to produce personal companions to not only assist their aging population, but also to interact with humans in human-like ways and expressions. (BTW, he cites in his book Love and Sex~ that there’s a Japanese firm that rents out sex dolls—Real Doll—to Japanese businessmen needing to stay overnight in a micro-hotel. So their concept of “companion” has more than one meaning.) Dr. Levy also sees promising trends in the application of AI to make robots emotionally sensitive and responsive to human partners. It takes only a little and plausible leap of faith that all of these trends can then combine into having almost the ‘perfect’, sensitive, and caring partner that someone would be attracted to. For emotional AI, Dr. Levy cites the ELIZA software as an early stage in this continuing development.

How about human acceptance of this concept? Dr. Levy notes that while initially this would appeal more to the physically deformed and/or the socially maladjusted, as robots’ AI matures to a much more sophisticated level of emotional empathy, it would become more acceptable to the more general population. He also notes that even initial moral and religious objections would be overcome citing historic parallels in inter-racial and same gender marriages. So Dr. Levy is pretty confident that machine/human relationships and marriages can take place by 2050.

While I think Dr. Levy’s book is an important seminal work on the subject, I think it also suffers from two major flaws:


  1. It depends on a favorable sequence of linear developments and events to achieve the 2050 goal let alone intimate robot/human relationships
  2. His relationship models seem based on the robot being the relationship partner that is the one that has to be subservient to the human partner.

As to the first objection, it’s always hard to make futurist technological projections—even for a knowledgeable AI person like Levy. The only successful one of recent times was Moore’s Law, but that dealt with the well-bounded field of VLSI technology. When one is dealing with multiple disciplines as is done in the book, it’s more like the embarrassingly discredited 1974 book, “The Jupiter Effect” predicting that a chain of events starting with planetary alignments would have led to massive California earthquakes in 1982 …didn’t happen …oops! With all the multiple disciplines that Levy tried to pull together, it’s almost certain that some of them would not go according to his estimates. Then of course there’s usually one or more out-of-the-blue events that change and/or accelerate everything. There are too, of course, retrograde trends that also historically happen.

Perhaps my biggest objection is the ‘subservience’ issue. If a silicon-based robotic lifeform is sophisticated enough in it’s AI capabilities to recognize and empathize with its human’s needs, wants, and desires, Levy seems not to allow that such lifeforms may achieve self-awareness. If ‘they’ ever got to that point, might not ‘they’ put their own ‘emotional’ needs at least on par with the humans. Revolt of Battlestar Galactica’s Cylons might then be possible. Relationship-wise for me, I prefer partners that choose to be with me rather than be ‘programmed’ or forced to be with/love me.

I do see a place for robotic lovers, however. I think that for a limited time period and given the proper ‘horniness’ disposition, it would be easy to have sex with a particularly ‘attractive’ robot lover—especially if there’d be no unfavorable consequences. Yes, this could be considered prostitution, but there would be at least the dissonant argument that “…it’s only a machine”. There would also be that variety to appeal to many “hot, zipless fucks”.

I heartily recommend reading Levy’s “Love and Sex with Robots: The Evolution of Human-Robot Relationships”. It’s an important book and presents at least an enticing blueprint of one way as to “how things could happen”.