Part
1.This was THEN….
In
the beginning, in our Milky Way galaxy, and in a rather miraculous
workaround, the earth, moon and sun convened at some point
to create life. And, eventually,there came a time when the
miracle of life and a sense of self was everything.
And
we knew it. But now….
This
knowing, of course, implies a mind alive not just to survive,
but to discover, and to create. Not surprisingly, while we
are the first and only species to argue about what is intelligence,
for us to modify that word with “artificial,”
or even the word “alien,” is to muddy the waters
of what it means to be human.
And
yet, here we are at an A.I. crossroads, puzzling over the
possibility of further evolution in the melding of minds with
machines (Part 2.).
Of
course, it is a special moment in history when we can opine
on the human-machine continuum and the creation of what might
be autonomous carbon-silicone entities.
While
we do not need to get into the weeds with A.I.,we can briefly
summarize. Traditional A.I. simply processes data and completes
tasks locked in by predetermined rules (algorithms). Generative
A.I. systems, trained on the fertile fields of data-harvesting
known as large language models (LLMs) like OpenAI’s
ChatGPT, are exceptionally gifted at analyzing vast amounts
of digital texts. But AGI (artificial general intelligence)
or AMI (artificial machine intelligence) is a next level promise
to improve itself with no human intervention and through its
enhanced “neural network” toexhibit through data-driven
probabilities (statistical inferences) a genius for pattern-recognition
and “self”-correction.And, lastly, all A.I. competing
countries are racing to be first to create ASI or Artificial
Superintelligence.
Furthermore,
to the extent we are what we make, A.I. relies on an exquisite
digital sense of what is real, and yetA.I., in fact, never
senses the fine quiddity of anything. Thus, if A.I. is an
avatar of ourselves then what A.S.I. (Artificial Superintelligence)
might be asking is not whose lives matter, but, rather, “whose
existence matters?”
And
considering that generative A.I. and AGI are technologies
designed to outsmart us and, perchance, to replace us, the
coming of A.S.I. will be a transformative moment in human
history that raises at least three key questions: Are we on
the verge of closing out the age of biological autonomy? Is
humanity at risk of engineering its own extinction? And lastly,
harsh as this may sound, is it possible that the future doesn’t
need us? Or framed another way, perhaps the presence of A.S.I.
is really asking “whose existence matters?” and
“what, if anything, should come next?”
While
all these questions are begging for answers, I suggest we
first provide some historical perspective, some sense of where
we’ve been. And we need to pull on two historical threads
that tie A.I. behavior to both (A) our more distant past as
well as (B) our more recent past.
(A) Crossing the African savanna around 4,000,000 years ago
were these 4 ft. tall bipedal hominids now known as australopithecines
(meaning southern apes—but more human than ape). They
had small brains measuring barely 450 cubic centimeters (cm³),
a bit smaller than a softball and about 1/3 the volume of
today’s human brain (1400 cm³).
Assuming
that an adaptive radiation occurred some 3,000,000 years ago
the result was that several similar, yet significantly different
hominid species, appeared and coexisted. From one of these
small-brained groups, around 2,000,000 years ago, our own
genus Homo evolved. Within that genus were three subsequent
human species: Homo habilis, Homo erectus, and finally Homo
sapiens.
The first of the human species is called Homo habilis (or
handy man), largely because of skills with stone tools; and
the evolutionary successor, Homo erectus (or upright man),
made his appearance in the Pleistocene period about 1,500,000
years ago. At this point, while retaining many simian features,
the size of the brain had reached at least 750 cm³.
Of course, encephalization (the increase in brain size relative
to body size) continued apace, so that approximately 250,000
years ago brain size had basically doubled to 1400 cm³
and Homo erectus gave way to the incredibly adaptable Homo
sapiens (modern man or us).
As for the sophistication of our brain, whether due to more
specialized tools, a carnivorous diet, social cooperation
in hunting game, or a spoken language, the reasons for the
brain’s growth remain uncertain and debatable. While
this evolving hominid brain and its neural explosion spanned
some millions of years, it is stunning to realize that today’s
A.I. technology – in just 75 to 100 years – has
produced a leap “akin to the intelligence” it
took nature eons to develop.
As
for (B), the second,more recenthistorical thread, we (somewhat
arbitrarily) begin in the fifth century B.C., with the Greek
philosopher Leucippus, who, maintaining his faith in a predictable
universe, stated with great confidence, “Nothing happens
at random but everything for a reason and by necessity.”
He believed that if certain phenomena appear to be unpredictable,
such as hurricanes, volcanos, eclipses, or the skipping passage
of a butterfly, it is only due to man’s inability to
decipher the combination of predictable forces causing them.
Really? One could only hope.
However, in 1687, the English polymath, Isaac Newton, outlined
these predictable forces in his Philosophia naturalis
principia mathematica, and attributed all movement to
three laws of motion. Here he was asserting the existence
of a mechanistic universe able to be measured and governed
entirely by cause and effect. The challenge, of course, was
finding that cause. Again, quite a reach and not so easy.
In the eighteenth century, the French mathematician Pierre-Simon,
Marquis de Laplace, was given to boasting that, given the
position and velocity of every particle in the universe, he
could then predict all movement for the rest of time. This,
too, was clearly wishful thinking.
As an example of the relentlessly unpredictable, in1827, the
Scottish botanist Robert Brown observed that when pollen grains
were suspended in a liquid, each grain (for no apparent reason)
appeared to move about randomly. This phenomenon, called appropriately
“Brownian motion,” remained a puzzle to be solved
by Albert Einstein.
Einstein explained that such pollen grains are pushed around
by the collisions of numerous water molecules. And since there
was no way to know the position and velocity of all the billions
of water molecules, one could never determine the next position
of a pollen grain. Such random behavior, unresponsive to those
Newtonian equations of motion, could only be described by
statistical probabilities, a topic neither Leucippus nor Newton
were familiar with.
Lastly, in the 1960s, as the father of chaos theory, the meteorologist
Edward N. Lorentz from M.I.T. observed that it is impossible
to know all the precise conditions necessary to predict the
weather, and the very flapping of a butterfly’s wings
might generate an air flow that eventually triggers a tornado.
This scientific hyperbole is sometimes referred to as the
“butterfly effect.”
With
all that in mind, consider the following: despite the microwave
satellite stations forming a computational membrane swaddling
the globe, and despite bingeing on massive digital internet
data, the task of ever collecting “enough” data
is impossible. Furthermore, the responses of any A.I. system
are an algorithmic attempt to simulate, in silicone, the carbon-based
connections between neurons in the human brain.
Given
these machine-learning algorithms, what generative A.I.’s
neural networks have achieved already (as with Google’s
Deep-Mind or China’s DeepSeek) seemingly matches or
exceeds the problem-solving ability of even our most gifted
humans.
To
be sure, one might suggest as a metaphor that A.I. could be
technology’s version of the fluttering of a butterfly’s
wings. And without hesitation, the tech wizards assert that
the results may not be either fully understandable or even
knowable. In other words, given the very real issue of our
survival, since A.I. has no moral compass, there is the compelling
and inconvenient question of how much autonomy should be given
to any such unpredictable silicon A.I. system.
Of
course, if we get this wrong, apocalyptically speaking, we
may not live to talk about it. And so, to that end we offer
the following:
We
cannot remember, but . . .
Beneath
the startling simmer of new stars
we never knew our cosmic dawn,
or the sun’s first solar winds
devouring the blackness
and splashing light over land.
Yet,
soon after, so to speak,
those earth winds
scouring paleocontinents
and surfing the Gaian tides
were urging micro-organisms to multiply,
to conceive and bear witness to life.
But
much, much later, after ancient winds
were no longer shepherding clouds
and pterodactyls,
and after tens-of-millions of full moons
lit their way through limbs and leaves,
those differing tribes of hominids
still evolving, were still moving
from then towards us.
And
despite time’s corrosive tyranny,
leaving only fossils behind,
we pressed into the future relentlessly,
wanting to know, needing to remember,
planning to control, believing we could,
while making, dreaming,
savoring that driving sense of survival,
coveting each second smitten with the possible
and forging the how of how to be human.
And
against the cave wall, as Plato wrote,
there was only the shadow of the real,
never the object itself
nor any name or metaphor so incandescent
as to burn through and capture
the very essence of anything as it is
or even a self.
Beyondthe data-sense of any fiber-optic A.I.,
wherein and to which meaning does not matter,
our very sentient self of flesh and blood
(sharing DNA of bone and branch,
of bird and tree) is ever rejoicing
in knowing that in every sunrise,
we are salmon, frog, oak, rose, hawk,
seed and egg,
singing of a lived past, a future purpose
and the driven needs of a mind pausing
to cast our dreams where hope prevails.
also
by R.J. Andres
Is
It More Than Just a Thought
Archaeopteris
and Us
Portrait
of Black Hole
Poetry
Two
Poems
