Re: Merging with machines inevitable, scientists say

Tim Tyler <seemysig@xxxxxxxxxxxxxx> wrote:
A recent article:

"Merging with machines inevitable, scientists say"

http://memebox.com/futureblogger/show/303-merging-with-machines-inevitabl
e-scientists-say

I seriously doubt it. The future is far more likely to be well-described
as a machine takeover, than a merger between man and machine.

I see the odds of the machines taking over being exceptionally slim.
There's just no way to justify how it could happen. Anything humans can
understand as being a threat will be treated as a threat and dealt with.
We put up guard rails to protect us from anything that shows itself to be
dangerous to us and AIs which are motivated to survive will be far more
dangerous to us than atomic bombs or global warming. As such, we will
build whatever defensive systems required to keep us safe. If we need to,
we will simply outlaw all forms of "AI".

For some reason, people think that "intelligence" and "survival" are
somehow related. They have nothing to do with one another. Intelligence
is the ability to find behaviors which maximize a goal and nothing else.
Humans are programmed by evolution with the goal of surviving. We won't
build AIs with the goal of surviving. Doing that would be as stupid as
giving atomic bombs to every child. Humans aren't that stupid and it won't
happen.

For the few humans that want to be anti social, we will invest in whatever
defense and protective systems are required to make sure the small
anti-social minority can't kill us with their AIs.

There's no path from a world full of smart machines taking care of us (just
like we already have today only with "smarter" machines), to a world where
the iPods and toasters rise up and take over. The entire ideas is just
idiotic outside of science fiction.

AI will outperform the human brain by an enormous factor.

The whole idea that today's crappy, evolved biotechnology will persist
for very long in the face of engineered solutions seems ridiculous.

It's not an issue of which is more advanced. It's an issue of what humans
will engineer. Why on earth would we every in a billion years engineer a
machine which would want to take over and control the allocation of energy
and raw material from us? Again, humans aren't that stupid. We are
motivated by the way evolution has built us to not let that stuff happen to
us and there's nothing about evolution that's going to program us to be
different.

On the issue of evolution favoring the life forms (aka systems) that are
most efficient at the dispersal of energy, what makes you think that a
machine could be more efficent at the utilization of energy than a human
directing a machine? What advantage does a machine have over a human
directing the machine?

For example, lets say some typical idiots thought god wanted them to build
machines to take over the world, and they set about building these highly
advanced AIs for the sole purpose of taking over control from the humans.
They create the 9/11 of AI by releasing these things on catching everyone
off guard. Next think you know, we have a real problem with some
unexpected attack by these AIs.

Still, no matter how smart the bad AIs are, the majority of energy and raw
material is still controlled by humans. And, since humans are motivated to
not lose control of that energy and raw material, mankind will throw as
much of it as required to stop the bad AIs. We will spend every resource
we have control over to stop the bad AIs - even if it means 90% of the
humans will die in the process. How can these bad AIs ever "win" when
humans still have control over the majority of the resources? The answer
is they can't.

The only way for bad AIs to win, is to take over control of the majority of
resources before the humans have realized they lost control. But if there
are AIs smart enough to do that, humans will fear that possibility worse
than we fear global warming, or the boggy man. As a result, we will build
advanced AI systems to monitor what is happening to make sure there are no
"bad AIs" which someone has created. We as a society will invest any
amount of resources required to make sure the danger never happens.

It makes no difference how "smart" the bad AIs are because no matter how
many "bad AIs" there are, we will always have more "good AIs" working for
us to protect us from the bad AIs. Humans don't need to be smarter than
the AIs any more than we need to be faster than a bullet to protect us from
bullets. We just need machines that give us the more power. The AIs we
build to protect us will be the machines that allow us to fend off any "bad
AIs" that happen to get build by idiots.

A machine take over just won't happen.

But what might happen?

In the short term, say the next few hundreds of years, nothing different is
going to happen. Machines which are already smarter and stronger than
humans, will get smarter and stronger leaving us with no mental or physical
advantage over the machines. This will cause some social changes as our
work-for-a-living model of society falls apart, but nothing will change
otherwise. Humans will still be in charge and the machines will still be
doing all the heavy lifting for us. The invention of AI is no more
significant to the advance of technology than the invention of the steam
engine. It's just another machine humans will use to do the heavy lifting
for us.

But in addition to AI, the far more complex issue of genetic and biological
engineering will confront us as we learn an endless number of ways to
modify humans. That will make us question what we are more so than the
invention of AI. For example, along with the invention if AI, lets say we
invent a brain implant that allows us to replace our brain with a machine
at birth. These humans with the brain implants act like humans in the most
case (the brain implants were carefully engineered to allow this), except
they have a few important advantages. For one, the brain state can be
dumped to storage for safety at any time, which allows, with the help of
DNA based body part cloning, for the person to be re-built in case of
accident. And, these types of humans will be able to live forever by
replacing the brain implant as required and by replacing the body parts
with cloned body parts as required.

When these sorts of changes become possible, what will humans allow to
happen? It's not the advancement of AI that is going to cause problems
because we will be able to build AIs that aren't humans and don't have
human desires and as such, there will be no problem with society knowing
"who" is in charge. We won't turn over the world, or our government to the
AIs.

But the more we can modify humans with implants and with genetic
modifications, the harder it will be to identify what exactly should be
allowed a place in society - which ultimately means answering the question
of who and what to give power to - the power to control the allocation of
energy and raw material.

As long as "man" is easy to separate from "everything which is not man"
society will stay much as it is now with humans in control. But when it
becomes impossible to define what a "human" is because the world becomes
filled with an endlessly complex mix of hybrids, then we will have the
general problem of society transforming into something controlled by the
strongest "things" that are around. In time this could well end up with a
transition from biological humans controlling the world to some totally
non-biological form in control.

But again, it's not a "takeover" by the AIs. It's a merging of biology
with engineering which then drifts slowly over time to a society of
something very different.

AI will be here shortly (50 to 100 years at the most), but this
transformation of humans into hybrids will take 1000's of years and without
knowing which types of hybrid technologies show up first, we can't hope to
make any predictions about what it will look like.

The change might be far more biological in nature. The real changes to
humans in the future might all happen at the cell level instead of working
with brain implants or other "bionic" replacements. We might genetically
start to re-engineer DNA and the cell and change man from the cell up.
Once we learn how the cell works and how all the DNA works to produce a
human, the first real changes might be from the bottom up by changing DNA
and by changing how systems in the cell works. We might end up with a cell
design that has an exponentially larger amount of DNA in it which allows us
fine grained control over all the systems the cell grows into. Why implant
a better brain into a human if we can change the DNA and the cell to build
us better biological brains? What if we come up with a biological brain
design that can dump all it's programming and transfer to a second brain?

There's no end to the stuff that might happen, and depending on what type
of technology develops first, will determine the path that human evolution
takes.

Humans aren't going to give up control to AIs, but they will use the
technologies they develop to alter the path of their evolution. Or more
accurately, the technologies we develop simply will alter our path of
evolution.

But it will be humans that are in charge, and it will be our complex
evolutionary path which will inherit the future, not the AIs.

Intelligence and the will to survive just don't have anything to do with
each other. Life and evolution is about systems with a will to survive.
Intelligence has no inherent will to survive and there's no reason we will
build AIs with a will to survive and a billion reasons why we will do
everything we can to make sure it never happens.

The creation of AI technology isn't going to change the evolution of man in
any interesting way. It's all the medical technologies that allow us to
re-engineer humans that will cause all the problems.

--
Curt Welch http://CurtWelch.Com/
curt@xxxxxxxx http://NewsReader.Com/
.

Relevant Pages

  • Re: What are we going to wish for?
    ... Corporations are already AIs. ... customers happy, and in the end, this makes all the individual humans ... long as we keep control of them - as long as we keep control of their ... of the government with our equal votes. ...
    (comp.ai.philosophy)
  • Re: What are we going to wish for?
    ... customers happy, and in the end, this makes all the individual humans ... But these giant and very smart and very powerful AIs only work for us as ... long as we keep control of them - as long as we keep control of their ... true world government. ...
    (comp.ai.philosophy)
  • Re: The wirehead problem
    ... controlled only by the will of humans. ... These industrial machines which are not called "robots" ... An robot with inteligence built to protect itself from harm will be a very ... I'm still in control of all this stuff ...
    (comp.ai.philosophy)
  • Re: Google AI: well at least its not not evil
    ... our society is organized around labor compensated consumers. ... Strong AI is going to kill capitalism as we know it - at least for humans. ... The AIs will continue to use it in their markets. ... everyone gets an equal slice of the production by the machines. ...
    (comp.ai.philosophy)
  • Re: Will AIs be religious?
    ... Once machines can learn by themselves, ... That's would be true only if you believed human personality or AI ... Actually, when the job is interacting with humans like by taking orders, we ... AIs with no long term memory might the the safer solution. ...
    (comp.ai.philosophy)