DTWolfwood said:
They don't "make" the next generation as how they see fit, they can only teach and pray they turn out the way previous generation wants them to be.
That's merely due to an inadequate degree of control of the "mechanisms" of human reproduction. That kind of control will (in all likelihood) be possible within the next few human generations, and then it will become
very relevant for all of us to sit down and decide how intelligent we want our children to be, and what behavioural tendencies we want to encourage or eliminate.
Biological processes are not necessarily any more random or complex than mechanical ones. They only appear so through lack of knowledge and control. By the time it becomes possible to built functioning machine intelligence, it will almost certainly be possible to exercise the same degree of control over the creation of living things, and that is not to say complete control. There will always come a point where chaos theory takes over, where the sheer complexity of the system you are dealing with produces an unexpected reaction, whether that system is organic or machine.
This clearly occured with the Geth. There came a point where the Geth ceased to be simple and predictable machines and became sufficiently complex in their operations to begin demonstrating understanding (by asking questions based on concepts and data they had not been programmed with). At this point, what the Quarians had wanted or had tried to precondition ceased to be relevant.
Tali actually explains why the Quarians wanted to shut down the Geth in ME1. It isn't because they didn't consider them alive, it's because they knew that intelligent beings would resent their position as slaves. Even the Quarians knew the Geth were more than machines at that point, their choice was based on simple xenophobia and panic.
DTWolfwood said:
Its the same as all those tiny robot experiments that have been done, the more stupid machines you link together, the smarter it seems their actions become. (Swarm Intelligence [http://en.wikipedia.org/wiki/Swarm_intelligence]) They are nonetheless artificial.
That's anthropomorphism, you've assumed that intelligence must resemble (individual) human intelligence in its function, not even in its output, in order to "count" as intelligence. Why?
At the end of the day it doesn't matter how something thinks. We don't ask whether minor neurological differences between humans make one capable of "real" intelligence and another only capable of unconsciously mimicking intelligence.
The only remotely logical argument against the Geth being intelligent is the "Chinese room" argument, that being able to produce symbolic data doesn't imply understanding. Yet the Geth clearly create and innovate on their own programmed data. They design their own starships and space stations, they conduct scientific research, they rewrite their own programming and evolve their own software to counteract threats, they come to opinions and conclusions which directly contradict their programmed role. They've broken the first axiom of the Chinese room, that programming is purely syntax without understanding and that all machines can do is move symbols around.
The only rational principle in this case is to assume intelligence, it doesn't matter whether the intelligence is "real", it matters that it is demonstrable.