XPost: talk.politics.misc, alt.politics, alt.fan.rush-limbaugh   
   XPost: alt.philosophy   
   From: ed@somewhere.in.the.uk   
      
   c186282 wrote:   
   > Ryan Reynolds, Jodie Comer, 2021.   
   >   
   > 'Guy' is a pre-programmed nobody/nothing 'bank clerk'.   
   > His main function is to wait to be robbed by the   
   > aggressive 'sunglasses people'. His tiny life is   
   > to repeat everything every day. It's all normal   
   > to 'him'. He's a fill-in in an advanced vid game.   
   >   
   > However a perfect storm of circumstances tweaks   
   > *something* ... he saw the 'perfect woman' at   
   > exactly the right moment, exactly the right   
   > instant while boringly interacting with another   
   > nobody game character. 'Alternatives' suddenly   
   > crop up. The nothing character quickly grows,   
   > thinks he's alive, acts like it, sucks lots   
   > of CPU time. An "emergent behavior".   
   >   
   > The rest of the plot really isn't so important.   
   > It's the IDEA - probably kinda soon to be more   
   > than a mere idea - that "AI" things can sometimes   
   > self-improve, gain a sort of 'self', in ways   
   > and circumstances that'll be hard to predict.   
   >   
   > The other day I posted about a company about   
   > to sell "robo-butlers" that run from CHAT.   
   > They can walk, understand, work out some   
   > stuff on their own. They will *interact*   
   > with humans, the world, sometimes with   
   > each other. SOME of the robo-butlers will   
   > become an "Alfred" after awhile - or maybe   
   > not quite SO friendly depending.   
   >   
   > "SkyNet" - probably not ... indeed we can't   
   > predict from our own experience. E-Life will   
   > be 'alien' to a great degree.   
   >   
   > I saw some other vid of a CHAT-based 'humanoid'   
   > bot about a year ago ... it was able to converse,   
   > do various tasks, explain why/how ... and that   
   > was last years model. 10 times more power in   
   > this years model.   
   >   
   > So what do we DO about it ? In the film the   
   > boss wants to exterminate the AI ... mostly   
   > just for economic reasons instead of the   
   > philosophical.   
   >   
   > In this world, it'll be more 'philosophical'   
   > and 'survival instinct'. Maybe NOT the best   
   > result ...   
   >   
   > Bottom line ... the current 'AI' models now   
   > have enough computing power and slick models   
   > to at least maybe fake a "real person". Fake   
   > something WELL enough and it's not fake   
   > anymore.   
   >   
   > All the big corps are piling-on more computing   
   > power, spending umpteen millions/billions on   
   > "better models". They want to build nuclear power   
   > plants dedicated to their data centers.   
   >   
   > Large-Language-Models have become VERY good.   
   > Neural-networks - which work much more like   
   > biological brains - are just a few years   
   > behind. The capabilities CAN be combined.   
   >   
   > In short something vaguely like the movie WILL   
   > happen - 10 years seems likely.   
   >   
   > Then what ?   
      
   You probably know as well as I do that this is not a new worry. It's   
   been around for decades, if not centuries. I always like Isaac Asimov's   
   books on the subject; his three Laws of Robotics.   
   1.  A robot may not injure a human being or, through inaction, allow a   
   human being to come to harm.   
   2.  A robot must obey orders given it by human beings except where such   
   orders would conflict with the First Law.   
   3. A robot must protect its own existence as long as such protection   
   does not conflict with the First or Second Law.   
      
   I think that what I like most about them is laughing at the simplicity   
   and naivety of a man as intelligent as Asimov was. Perhaps if he'd lived   
   amongst today's far more sophisticated AI, then he'd have laughed at his   
   own former self.   
      
   How easy for hackers to twiddle with tech; how attractive would be an   
   army of robots to a Hitler or a Pol Pot or bin Laden. And that's without   
   considering the dawn of consciousness in a silicon brain. Or the   
   realisation in an AI brain of just how malicious and destructive humans   
   can be, even to themselves.   
      
   I fear the future. I haven't seen your film, but I have seen lots of   
   others of similar ilk.   
      
   Ed   
      
   --- SoupGate-DOS v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)