If merging ourselves with computers to produce a new kind of human is BETTER with respect to evolutionary requirements then if it's given half a
chance it -will- happen. Just like machines did. Just like the stone age moved to the bronze age. Just like people moved from sails to engines. It
goes on and on.
If it makes us better then how is it bad???? This competition isn't being judged by humans, but by nature. We don't know what the end results will be,
only nature does. Just do what you think is best.
I like to program as a hobby. Lately, I've been wondering why I don't learn more about multi-processors and programming for them. The reason I say
this is because the brain is a parallel processor, much the same as multi-processor systems. (When companies bumped into the overheating limits of
chips, they started making multiple cores to compensate.) And programming for multiple cores is different than programming for one. It seems to me
that learning about this is very worthwhile.
Parallel processing means the brain does multiple things at the same time. This doesn't mean there're not some sequential processes. Some processes
will require other processes to precede them, even in the human brain. I read that images we see are broken down into these 4 conditions, in the
brain: color, shape, motion, depth. These're processed separately and in parallel. However, in order to make a full cognitive assessment about what
you see you must first wait for the brain to finish these processes. It might be that we can make decisions even with limited information - so maybe
we can examine the information as it's being processed. But a reliable conclusion cannot be made too soon.
I do not know how far away we're from achieving the same instructions-per-second capability of the brain. The reason I say this is because I do not
believe we understand the brain well enough to judge its true instructions-per-second rating. In the computer world, instructions on different cpu's
can operate at different performances because of differences in microarchitecture. It depends on the instructions you're using too. So just getting a
overall instructions-per-second estimate of the brain is not helpful until we can understand what sorts of instructions the brain is using and how
well they can translate to modern processors. For all we know, we will need to rethink cpu's to imitate the human brain.
There's no doubt that -right now- the human brain does some things better while computers can do other things better. The brain is excellent at
analyzing patterns and abstract things. Computers are not as good at that. They're infants. But computers are excellent at mathematics. They're so
good that they can perform in a hour what would require all the worlds scientists a year to do, or more. Whether this stays true in the future is
unclear. It could be that even if we did accomplish synthetic brains on a chip we would still find that they suck at mathematics and other linear
processes. Maybe intelligence algorithms and architecture are just fundamentally different from non-intelligent instruction designs, so there'll
always be a divide.
I do believe that all of the worlds computers probably are doing more instructions per second than a human brain, but the type of instructions is
what's so critical. Maybe we're underestimating biology.
To think that all of that processing is going on in something you can hold in your hand, about 20 watts? Amazing. And so much of it's unclear. We take
everything that the brain produces for granted.
I still see AI as complimentary. We will merge with computers, not be destroyed by them.
To me it's an exciting future. It's what makes life worth living.
I see quantum computing the same way. It's complimentary. Not earth shattering.
The mistake people make is thinking that these things, once understood, become godlike. I don't think that's how it works. I think things work
together in nature. They augment each other's strengths. For example, if our human mind could exist entirely within a computer then it would be
possible to beam ourselves to the moon at light speed. This would be very useful for purposes of travel. However, once we get there we would need to
be able to interact with things. We can't do that if we only exist as information. So we must have a body. And I think that we cannot rule out that
biology might be found to have a advantage over pure robotics. If robotics is better, though, it will be used more. But maybe the logic behind biology
is its ability to repair itself. Maybe biology is better at that than robotics could ever be. Billions of years of evolution cannot be easily beaten.
Maybe there're physical laws that restrict it too.
edit on 16-3-2012 by jonnywhite because: (no reason given)