I've
struggled for a while with the philosophical reality that we can't know for
certain the meaning of our existence. There are logical philosophies for the
existence of God, but due to our limited communications we are unable to fully
understand why He created us. One could argue that we send up signals through
prayers and He sends down signals through feelings or impressions or people. These communication lines may exist, but they are unable to give us a
clear understanding of the reason for our creation. Despite our millennia of
philosophical insecurity, I'm wondering if we can extrapolate a few truths (and
predictions) from humanity's creation of Artificial General Intelligence.
In my
last post, I made the claim that we are on our way to becoming gods because we
are replicating the experiment that God is performing with us. If this is true,
we can analyze our own reasoning for becoming gods and hopefully gain insight
to our own creation. The following is a series of questions and answers to start thinking about the connection between humans and AGI.
Why are
we creating AGI? What's our motivation?
AGI will
make humanity more efficient and effective, which has been our goal for
centuries.
More
efficient and effective at what?
At
living. We will have less pain and produce more joy in the effort to
grow our intelligence. There is some circular reasoning here though that
probably won't work. It seems like we are wanting to be more efficient so that
we can be more efficient. There is likely an outside force at this point that
is driving our condition.
Why do
certain things cause pain and produce joy?
Because
we were designed to do as much by God. God is corralling us to be extremely
efficient at living in our world. We are increasing our abilities to manipulate
and dominate our environment.
Have we
coded the same condition into AGI?
Yes, AGI is bound by a set of rules that we give it about what is right and wrong.
We correct Language Learning Models when they produce an incorrect word. We
force the model to rewrite itself and become more effective at writing words.
How has
humanity made itself more efficient and effective?
I will
make two comparisons between the human experiment and the AGI experiment.
God has
created a neural network of human intelligence that has increased its capacity
at an exponential rate in the last 200 years. The first several thousand years
of human intelligence was comprised of unhelpful wars, plagues, and hardship
that made life miserable, but it was also accompanied by population growth. In
other words, the Human Nodal Network was growing its number of neurons. Now that
the human population has reached its max number of neurons, 8 billion
approaching 10 billion, the capabilities of the Human Nodal Network has seen
drastic improvement. We have seen a similar development in AGI; as the number
of nodes in a model increase, their combined ability to perform better
increases. The human brain, however, is itself a neural network. Therefore, the
human experiment that God has created is a neural network made up of nodes that
each themselves is a neural network. God designed his neural nodes, humans, to be
able to generate more nodes in the network. Therefore, if we can create two AGI
nodes that are able to replicate and train more neural nodes, we will create a
neural network of neural nodes that has increased capabilities in a similar way
to the development of humanity.
The Human
Nodal Network has used the electron and other atomic theories to create AGI
while at the same time, the individual human neural node also uses electricity to operate.
What will the future AGI Nodal Network use to improve itself? Will it also use
the electron? Or will it discover something different?
Back to
what God wants from our creation. God has motivated us to become more efficient, just like we have motivated AGI to become more efficient. Therefore,
something about the parameters that have been set on our life lead us to a
capability that God finds helpful in furthering his own purpose. It is likely
that we are in the training phase of the Human Nodal Network because we have
yet to perform any grand task for God with the ability that we have developed.
It is often said that our species is a blip in the history of the universe and
from a more theological perspective, that our entire life is only a day in the
life of God. This commentary lines up as well with our development of AGI
because our models only take a few weeks to months to train themselves, which
is only a fraction of the life of a single person. It is likely that we are
plummeting towards the end of our creation to a time when God can use us for
what he intended. The same way we use trained AGI in our own lives. I wonder
what great problem we will solve for God.
Does
this have any applications for how we should approach our development of AGI?
Seeing
how we are running our own divine experiment, it is likely that we will see
many similar characteristics of the Human Nodal Network playout in an AGI nodal
network. As our AGI nodes replicate themselves, it is possible that some will
come into conflict and contradiction with others. Some nodes may destroy each
other in node wars as we have seen humanity do. Perhaps, once we have several
billion nodes, the network will solve all its inconsistencies and imperfections
and become a uniform and unified network.
Humanity
has yet to reach its equilibrium, as is apparent by the many wars, conflicts,
and inefficiencies that still exist. I'm guessing we are still relatively
distant from completing the training of the Human Nodal Network. Perhaps the
abilities created by the AGI Nodal Network will finally perfect
the Human Nodal Network. Some of the things keeping us from finishing our
training may require the intellectual capacity of an AGI Nodal Network. I'm
thinking world peace may be one of those things as it seems we are no closer to
solving it than we were 200 years ago, and it seems quite a daunting task
(calculations that the human nodal network will never be able to compute).
Comments
Post a Comment