Book Challenges Fearful Predictions of Machine Domination of Humanity

As machines get smarter and smarter, the pundits rush in to scare us with their idea of what the future world will be like.  Generally, it’s not a pretty picture they see.  There seem to be only two possibilities.  Either machines will annihilate the entire human race in an unearthly blast of shock and awe.  Or we will all be gradually dehumanized into a race of zombies, mechanically following every wish of the new masters.  “If we’re lucky, they’ll keep us around for pets,” goes the dark humor.

A new book now challenges that dreary view.  “The Virtual Librarian: A Tale of Alternative Realities,” a visionary science novel by Ted and Bob Rockwell, suggests another outcome: the machines will teach us the importance of our being completely human, so that the machines can fully carry out their necessary and complementary mechanical functions.

The book is, first of all, a short, lighthearted story, enjoyable just as a story, set in the very near future.   Its first surprise is that it presents a wholly different kind of virtual world from those historically and currently in vogue. The virtual worlds we read about, starting with the shoot-em-up arcade games and going through Dungeons & Dragons, World of Warcraft, and Second Life, let participants represent themselves with an on-screen effigy called an “avatar.” 

These can be created in an endless variety of forms and costumes.  But, functionally they are ventriloquist dummies, all of whose words and ideas are supplied by their human sponsor.  The virtual librarian, by contrast, is entirely software, like the voice you talk with when you call to see if your plane is late.

  This virtual librarian, called Lib, has two advantages over current models.  First, she talks with computer-synthesized speech, rather than selecting among pre-recorded bits of human talk.  This is presumed (optimistically) by the authors to enable her to escape the limitations imposed by pre-recorded phrases.  Second, her operating software is designed to be evolutionary; that is, to change its procedures as it learns from operation.  Her designers are slow to realize that complex, interactive systems with evolutionary operating procedures inherently become unpredictable and are uncontrollable from outside.

The story inevitably encourages the reader to think about the changing role of humans and machines (such as Lib), and has Lib patiently teaching her human operators that both she and they cannot act like machines and expect satisfactory results.  This situation does not arise out of some sort of mystical or spiritual break-though among the humans.  Rather, it is a logical consequence of the machine trying to optimize its performance and being blocked by the humans unknowingly failing to fully carry out their role.

Such a situation is not unknown in the real world.  Stanislaw Ulam, one of the pioneers in designing chess-playing software, tells of a test run of one of his early programs.  The program was designed to stop the game when one player wins or when a draw is reached.  But no instructions were given for when a checkmate is obtained, where playable moves—but no winnable moves—remain.   Ulam’s game had reached checkmate, and the program was frantically looking for a winning move.  Suddenly a new pawn appeared on the board, and started to find moves that might eventually make it a queen.  Lib found herself similarly stymied, so, like Ulam’s software, she started to invent unorthodox moves to meet programmed goals.  

Is that the first spark of consciousness?  Is that how a brain creates a mind?

Comments

[viagra online]

This book really scares me, all he said is true.

[MBT Online]

It's so nice to have you do all of the research for us. It makes our decision making so much easier!! Thanks.