XPost: rec.arts.sf.written, rec.arts.sf.science   
   From: droleary@8usenet2013.subsume.com   
      
   In article ,   
    David DeLaney  wrote:   
      
   > On 2014-01-24, Doc O'Leary  wrote:   
   > > As I have posted in the AI newsgroups, you can easily set up a standard   
   > > for an AI driver.  Run this experiment: allow a person to drive a car   
   > > remotely using only a video camera, measuring their performance as you   
   > > reduce the resolution of said camera.  Then you pick a "minimum   
   > > acceptable" point for their ability to navigate the environment.  That's   
   > > your Turing Test for AI drivers, getting around with only a crappy web   
   > > cam as your input.   
   > >   
   > > Compare that to how Google and others are solving the problem and it is   
   > > easy to state unequivocally that AI is *not* being used.   
   >   
   > And I hope you were roundly mocked in the AI newsgroups as well?   
      
   Not as much as a critical discussion would warrant.  The problem with   
   most AI "research" these days is that they don't really *want* to do AI,   
   they just want to get patted on the back for getting results with their   
   little cheats.   
      
   > Why _prevent_   
   > an AI from using anything other than the puny human sensorium as input?   
      
   Because the issue is intelligence, not a plurality of senses.  Are you   
   also one of those people who would also claim that people with fewer   
   senses *must* be less intelligent?  You sound like someone who would   
   fall into the long-discredited behaviorist camp.  You don't   
   fundamentally understand what intelligence is if you look no further   
   than sensory input.   
      
   > What's   
   > your beef with radar, sonar, ultrasonics, microwaves, gyroscopic sensors   
   > (humans have fluidic similar things but they're not based on the same   
   > principle), electric-field sensors, radio, or other input channels?   
      
   My "beef" is merely that intelligence is what happens *after* an input   
   of *any* kind is received.  Adding more input might make certain   
   processing easier, but it might also complicate things as well.  The   
   burden is on you to show you're on the path to intelligence.   
      
   Or, to put a finer point on how foolish your argument is, I challenge   
   you to show why it is that *you* should be considered intelligent enough   
   to drive when you don't have radar, sonar, ultrasonics, microwaves,   
   gyroscopic sensors, electric-field sensors, radio, or other input   
   channels.   
      
   > Show us an AI that thinks as well AS a human but not LIKE a human, because   
   the   
   > latter's gonna be a great deal more difficult.   
      
   No, it isn't.  Not to someone who has a good definition of intelligence.   
   Without getting too deep into it in this newsgroup, I'll just say that   
   one non-brain implementation of intelligence is evolution.  Feel free to   
   go elsewhere to explore further (may I suggest comp.ai.philosophy?), or   
   just keep posting here if you want to assert with confidence that you're   
   still right despite what reality shows.   
      
   --   
   iPhone apps that matter:    http://appstore.subsume.com/   
   My personal UDP list: 127.0.0.1, localhost, googlegroups.com, theremailer.net,   
       and probably your server, too.   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)