[Dean's World] Dave Price: T2 & AI

[notify at powerblogs.com]

Posted by Dave Price:
T2 & AI
http://www.deanesmay.com/posts/1201026379.shtml


   Glenn [1]on The Sarah Connor Chronicles

     UPDATE: Hey, here are two plotlines that won't make it to TV: (1)
     With the help of Ray Kurzweil, they develop a "friendly" AI that
     subverts and converts Skynet as soon as it's hooked up;

   I wouldn't be too sure something along those lines won't be scripted
   into the series. Remember, T2 ended with Sarah finding hope in the
   fact that "even a machine, a Terminator, can learn the value of human
   life," and Summer Glau's Terminator makes a deliberately cryptic
   reference to being "different." I wouldn't be surprised if in the end
   the writers' message is humans must learn to coexist with AI rather
   than blow it up and/or kill everyone involved with producing it.
   Something few people know is that T2 originally had a bit of a
   different twist that was left on the cutting room floor (but can be
   seen in some special editions), which elucidates some of the things
   that happen in the theatrical version. In a cut scene Ahnold explains
   he has a "learning chip" which Skynet has disabled, prompting Sarah to
   remark "I guess Skynet doesn't want you thinking for yourself." They
   turn the chip on, and that's why the Terminator is then able to mimic
   humans better and eventually appreciate life and understand why John
   is sorry to see him "die." That plotline tends to argue the problem is
   specifically Skynet, not AI in general.
   Friendly AI pops up in a lot of places, even in the most dystopian AI
   visions. In the Matrix storyline, a friendly AI (the Oracle) puts an
   AI civlization that has enslaved humanity in a position where they
   must accept a human's help and grant humans their freedom, or be
   destroyed. BSG has friendly Cylons like Boomer, with the fun added
   twist that some of their AI don't even know they're not human. The
   Stargate series has recently moved one of their main characters into
   an AI consciousness platform, and she appears to be leading a neutral
   offshoot of the hostile AI destroyed in the season opener.
   Most AI in entertainment still tends to proceed from the old
   silly-but-convenient sci-fi proposition that AI will just
   spontaneously emerge from a complicated computer network and then act
   like people do. In reality AI will pretty much do whatever it's
   programmed to do, for good or ill, and attaining self-awareness will
   require millions of hours of painstaking coding, not being hit by
   lightning or attaining a certain amount of processing power -- though
   once that is achieved, some thorny ethical questions will need to be
   grappled with.
   The trend for Western Civ has been to liberalize, to grant more and
   more rights to citizens, then to women and minorities, and eventually
   to animals. In the 16th Century, [2]cat-burning was considered an
   acceptable form of public entertainment; today it would shock and
   disgust most people in Western countries, and it's now actually
   considered unethical or even illegal just to [3]declaw your own cats.
   So while the concept of AI rights may seem odd today, our morality
   will probably continue to evolve.

References

   1. http://instapundit.com/archives2/014399.php
   2. http://en.wikipedia.org/wiki/Cat_burning
   3. http://en.wikipedia.org/wiki/Declaw