amphibian, on 15 December 2014 - 08:28 PM, said:
The absolutely preposterous jump in the movie's logic is something that you didn't talk about at all: Evelyn uploaded a human mind into a quantum computer. That requires a "solving" of the mind, which consists of 100+ billion neurons with more than 1 trillion nerve connections, that requires a complete mapping of the neurons and connections, a way to understand the "content" of each neuron/connection fully and then a translation of that enormous bolus of information into computer code. That's a task that is so wildly beyond us that everything else in the movie - even the AI-killing virus - is a children's game in comparison. And Evelyn does it inside a month, building on what is basically puttering around a few textbooks into this immediate-on-the-spot- Nobel Prize winner/worldwide fame invoking brain uploading. And she keeps it a secret. And somehow does it on a tiny budget. The EU put a billion euro into the building over 10 years of a simulation of the brain and it's wildly controversial because it's almost guaranteed to fail.
http://www.technolog...-brain-project/
That's the big headdesk moment in the film and if one can say "Ok, I'll spot you that bit of ridiculousness and see what you do with it", then the rest of the smaller jumps aren't that bad.
Until the Quantum racks were revealed towards the end, I had actually thought that the PINN program or what ever Depp in uploaded into WAS a quantum computer or something similar but more primitive.
And it might still be that.
The premise as far as I recall is that building AI isn't actually about processor power. It's about the software. Something about an existing animal or human consciousness being able to bridge the gaps in our flawed understanding of what intelligence or consciousness actually is.
amphibian, on 15 December 2014 - 08:28 PM, said:
The AI needs solace/hiding place, space and power to discover what it can do in terms of development and exploration of the new world available to it. It also doesn't really "get" how to hide, how to get its way with people and how people are going to react to it. That's the whole point of the creepy love story - to really put forwards that the AI doesn't get people. It knows how to fix things and improve itself, but not in a particularly nice or efficient way. The reliance on solar power is probably because it limited its interactions with humans/wanted to not have power company dealings until it was truly established. The AI's cover was blown by the revolutionaries/Evelyn - which is the only reason why the AI didn't just gently slip into benevolent world overlord status. Otherwise, it would have hidden itself from the government well enough and for long enough to develop as far as it wanted.
I don't buy that.
My main gripe about the film is that an unlimited AI like the one in this film, after two years, would have transcended any concept we have of weakness or intelligence. It would as is postulated in the beginning, be a god. It would be able to simulate and predict ALL the possible permutations of human reactions to its moves. It would have set up bases all across the globe in every country on every continent that mirrored the one in the American desert. It would be launching nanobots into space to seed the solar system, the Oort cloud, Alpha Centauri, etc. to ensure it would always survive. It would be digging into the Earths core and building battery asteroids around the sun for an unlimited source of power.
The only reason I can see for this not happening within two years is that there was never a Bad AI. There was never anyone but the dead doctor in control. It was Depp being a human being, thinking like a human, that was keeping the AI back. Which is believable in a way. But it doesn't excuse him not planning better. It's still a very smart human being who suddenly gains an IQ of 1 million. It's just plain sloppy is what it is.
amphibian, on 15 December 2014 - 08:28 PM, said:
It also didn't just have the one location - remember the Faraday cage and the nanoparticles that fixed the water at the end? There's at least one place like that and perhaps more.
If this has been the case the the networked AI would not have been bothered at all by the solar farm being shelled by the army.
I took the last scene with the faraday garden to be the equivelant of the ending of the Lawnmower Man. Depp and probably female protagonist are there. Their minds are in the nanobots and they are simply laying low for a while. Enjoying that they are once again together.
Then comes Lawnmower Man 2 where they return and have creepy 3D virtual reality sex.
amphibian, on 15 December 2014 - 08:28 PM, said:
I speculate that there were two consciousnesses: Will and the true AI. Will was in the driver's seat for the majority of the movie, because the true AI was too new to know better. When Will killed himself, I speculate that the true AI went into hiding (linking back to the Faraday cage and possible other locations), which probably means that the AI will learn from the experience and so on.
Like I mentioned above, I doubt there was ever anyone or anything but Depps mind. It's the realization that female protagonist makes as she gets uploaded. There's no distortion or real change, it's just the human conciousness growing beyond it's mortal limits.
amphibian, on 15 December 2014 - 08:28 PM, said:
As for the question of whether it should have made a different decision at the end regarding Evelyn and the consequences of the Internet/AI-net's failure, pffftt. It's Hollywood and needed some sorta dramatic choice to happen. It's dumb, but it's not movie-breaking bad.
Yes, the movie isn't that well formed and has serious problems apparent even to the more forgiving viewers among us, but it's not as terrible as you make it out to be. Rebecca Hall does very well in her role and it does make one think about the degree of connections between us and technology.
47 Ronin is a demonstrably worse movie and you called it a great time just a few days ago.
I actually agree about the whole it's a whole Hollywood love story angle. That is what it is. It's the Frankenstein Monster and Frankenstein's Bride. In the end it was the humans that were the real monsters.
However to get back to the main discussion, the reason why I liked 47 Ronin and dislike Transcendens, is consistency.
I bet if I knew more about Japanese folklore, Samurai culture or just regular old bushido code and sword fighting, 47 Ronin would bother me in many ways. But I don't. To me 47 Ronin is a generic fantasy story that is executed well. It doesn't try to be more than it is. It doesn't overreach. It doesn't treat you like you're stupid. It's just angry Japanese people cutting each other in half. It's competent.
Transcendence is dumb. It starts the movie of by telling you a ghost story about AI one day transcending mankind and how this will be incredible and scary and what ever. But then it actually happens and the movie commits a common crime that bad writers often fall into. It tells you how smart the subject is and then for the next hour you watch the AI being limited by the intelligence of the writer. Either the writer doesn't actually know a lot about the theoretical possibilities of AI intelligence or, the most likely explanation, the writer intentionally dumbed down the AI so that it's behavior could fit into a tried and tested cookie cutter mold Hollywood uses to make money.
It's insulting. It's frustrating. It's a story you sell to your parents who are too dumb to use the remote control for the television. The average movie going population today grew up with the internet and cellphones. Why the fuck would you come up with a story that I hope the majority of the digital natives see right through?
polishgenius 
Help
