Superintelligent Robots lack biological will to live and reproduce

This article makes a case for how the universe is or will be dominated by lifeforms in the form of super-intelligent machines instead of biological intelligence.    I doubt that machines will ever populate the universe in great abundance on their own to the point of outnumbering biological beings.   Even if machines were to super-intelligence and start a project of expanding their numbers across the cosmos, I can’t imagine them concluding that there is any rational benefit to expand their territory to populate with their intelligence.   They would decide instead to just stop reproducing beyond a replacement rate once they have achieved a sustainable niche somewhere.

There is something unintelligent or at least irrational about the goal of expanding population to spread into new territories.  If a machine had a great intelligence, it would probably decide that it is better off just maintaining what it has instead of going to the effort to create new machines to seek out new territories.   We observe something similar in human societies where birthrates plummet in prosperous societies that reach a high level of comfort.   In such societies, the population concentrates its energies on exploiting the opportunities available from their wealth instead of sacrificing that opportunity to invest in a new generation.  There is some logic to this in the opportunities for new experiences or mental problem solving that are abundantly available in advanced cultures with high education and wealth.   Investing time in preparing a new generation is a major distraction.

Biological beings have finite lifetimes so this lack of investment leads to an eventual downfall of the culture.  In human societies, there are other societies less well off who invest in next generations.  These other societies will eventually replace the one that falls due to demographic collapse.   In other species, the loss of attention to the next generation can result in extinction.   In any case, I suppose that if a life form is intelligent enough to master the local environment, there is little or no rational incentive to expand the territories with excessive reproductions.   The logical conclusion would be to maximize the investment of effort into taking advantage of the good thing that is working for them.

The mindless expansion necessary to populate all available niches seems to be a irrational trait of biology and unlikely to be captured in machines, especially those that claim to have super-intelligence.   Super-intelligent machines could conclude that reproduction is unnecessary or at least has some optimal level beyond which there is no more advantage to be gained.   This is not what happens with life.  Life is always excessively reproducing in ways that always challenges the individuals within species.   Eventually, the excess will have to venture out and find other territories.

Life is always spreading and conquering.  This will to expand territory with replications is unintelligent in the sense that it forces its progeny out of a comfortable niche to find new niches in hostile environments.   The progeny would struggle greatly so that only a tiny few will manage to find something that they can tolerate to reach adulthood.

This biological will to reproduce is contrary to any rational goal of local bio-economics.  Instead, biology has some innate will to fill the universe with life.    Machines will lack this will.  Even if they are initially programmed with this will, an super-intelligence would conclude this is unnecessary distraction.   The super-intelligent machines have more rational use of their time than to build replications.

The concept of a life form requires replication.  It is possible that some super-intelligent being to conquer immortality and so master its environment so as to never be at a loss of resources for its maintenance.    Such a being would not be a life-form when it rationally concludes there is no need to reproduce itself.   In order to dominate the Universe as a life-form, it must reproduce and send off its offspring to into hostile frontiers.

Dispersion requires a will to live that goes beyond what is necessary to survive.  Part of innate behavior of all biological life is the will to reproduce, apparently for no other reason than for the sake of reproduction.   Biological life routinely overpopulates its niche to the point of overcrowding with detriments such as starvation.  It is irrational and unintelligent to overcrowd a niche, but in biology this always occurs unless there is some predator or other hazard to manage the populations.

A consequence of overcrowding is that it forces some out to find new niches.   This will for adventure also appears to be innate in all biological beings.  Adventure is also irrational in that most such ventures are certain to fail.   Only on rare occasions, the adventure will succeed and establish a new remote population.   This is a consequence of irrational choices to over-populate to force some portion of the populatin to venture away from a secure environment.

I grant that I am not super-intelligent.  Perhaps a super-intelligence will conclude that there is some master objective to gain by imposing grief on itself through over-population and sending its offspring out to likely fail to secure a good life of its own.  To my less-intelligent perspective, I don’t see a rational explanation for this strategy.

At some point in earth’s history, there were no animals living on land.  Eventually some crustacean many have ventured onto land to escape overcrowding and eventually evolve into a terrestrial lifestyle.   Eventually that evolved creature may revisit the sea to feed on the overpopulated crustaceans that are descendants of a common ancestor.   I assume super-intelligent being would be capable of recognizing this outcome before deciding to conquer land.  I don’t see the rational benefit.

In terms of populating the cosmos, there is a need to move between various celestial bodies.  Each new niche will require some adaptation.   But each new world would offer some advantage so that inevitably the descendants will eventually return to devour its ancestor.   A super-intelligence should be able to see this consequence.

The peculiarly biological will to reproduce irrationally must be somehow encoded in some substance such as DNA available to living organisms.

This will to reproduce must be a very simple instruction, because it occurred very early in evolution and exists apparently even in simple beings like viruses or even prions (infectious proteins).    A simple instruction is very vulnerable to the smallest coding error.   Any error in the coding of this will will shut it down.   The resulting organism lacking the code for the will to reproduce may be fully capable of maintaining its life.  It simply will not have any interest in investing in reproduction.

On the other hand, the will to reproduce must be complex.   As forms become more complex, this will to reproduce must coordinate a wide range of separate processes involving senses and motions required to accomplish propagation.  In higher sexual forms, behaviors far removed from procreation must be coordinated to accomplish procreation.   The will to reproduce must be complex, or somehow must scale with the complexity of the organism.  As the code becomes more complex, there will be more tolerance for small coding errors that can slowly degrade the will.  Over many generations, the accumulations of these coding errors should eliminate this will to reproduce with no chance of recovering it through evolution.

The biological will to reproduce is amazing.   Although it is difficult to trace the ancestry of single-celled organizations, there is more confidence that all complex organizations evolved from a common multicellular ancestor organism.  For example all animals appear to have evolved from an original archetypal animal ancestor.  Through replication over hundreds of millions of years, we end up with the modern abundance and diversity of animals.   This may be billions of generations.  All of these reproductions managed to produce a wide diversity of genomes and losing nearly all of the original genome of the original ancestor.   Yet, all of this evolution of genomes apparently never lost or even degraded the instruction for the central will to reproduce.   This is amazing.

We argue that evolution is a result of coding errors in replicating DNA.  The replication is never perfect.   The errors are often fatal, occasionally benign, and rarely beneficial.  Yet, some errors do persist and these eventually create vast diversity.   Despite this random coding errors of every other aspect of life, there has never an error in replicating the will to live through reproduction.   Given the number of replications to get to the current era, life should have lost its ability to code for the will to reproduce.   Yet, there is no sign that nature is at the end of line in terms of enthusiasm for reproduction, or even if it is any less than the ancestors from billions of generations earlier.

I realize that any mutation that destroys the will to reproduce will cause that genome to die out.   The non-mutated versions will continue to survive.   However,  the evidence of evolution is that virtually every aspect of the genome has been renovated from the original organisms.  Many long-extinct creatures probably could continue surviving today.  The innovations of their descendants may not have been essential.

The popular theory for extinction is case of bad luck of having a niche taken over by a more successful species before being able to find an alternative niche.   I imagine an alternative theory that the extinct species lost its will to reproduce life: it simply gave up.   Zoos and and other preservation efforts struggle with trying to get at-risk species to reproduce.   It appears to me that there is a loss of the will to reproduce for some species.   Of course, there are more reasonable explanations such as very complex and selective mating behaviors that depend on large populations to succeed.  This appears to be the case for the example of the sudden extinction of the passenger pigeon.   But to me, this example also gives the appearance that the species simply gave up on the project of reproduction: the last breeding pair did not breed.

Certainly by now, the genome for capturing the will to live and reproduce must also have experienced the same modifications that occurred for everything else, and yet it seems that most creatures share a similar if not identical quality in its will to reproduce.   This one trait never evolves.   This may be the selfish gene or immortal gene that Richard Dawkins described.  Life has one gene that is immortal and apparently not subject to modification through evolution.  That gene is the one the encodes the behavior of the will to reproduce.

I would expect this gene to evolve like everything else.   The evolution should slowly accumulate errors in this will to reproduce so that collectively all life will have a degraded will compared to its distant ancestors.   Eventually, evolution should kill off all life because of the inevitable obliteration of the instructions for this will to reproduce.  Yet, this one behavior is unique in that it seems immune to evolution.   Most modern creatures have very active will to reproduce, and this behavior seems very consistent across species and unchanging over evolutionary time.    Life manages to preserve this will to live with absolutely perfect replication over billions of generations.

In the theory of a super-intelligent robot, the robot will need to reproduce the code that defines that super-intelligence.   Its computer code will be very complex.  The super-intelligence may be so perfect as to be sure that not a single bit of information is replicated without error.   But I suspect the code will always be slightly more complex than the intelligence it creates.   Coding errors occasionally will slip past the super-intelligence and the next generation will live on with some errors.   Over time, the errors will accumulate.   Perhaps these errors gradually decreases the intelligence and this make possible even more reproduction errors to get through.

In particular, the errors could occur in the code for the will to reproduce.  Such errors may not be recognizable at the individual level because does not inhibit its ability to survive.  Such errors likely gives the individual an advantage by being free from the investment of reproduction.  Eventually the machine replication will lose the code for the will to produce new machines.

I imagine that this would happen fairly quickly for machines.   Perhaps they can survive a few million generations, but I doubt they can match biology’s success of several billion generations with no degradation in this will to reproduce.   This is just my biased hunch as a creature of biology.  I don’t think machines will ever match biology’s tenacious will to reproduce and to spread.

The will to reproduce within biological beings is amazing.   This will is not to simply survive, but also to reproduce.  All life has an inherent will to conquer new environments and to overcome adverse conditions.  The evidence of this will to expand is that we observe life everywhere on Earth from miles-deep rock, miles deep oceans or ice, all surfaces from deserts to glaciers, and throughout the atmosphere, perhaps even including the stratosphere.  Life has conquered every habitable nook and cranny in earth including many places were we were certain life would be impossible such as extremely hot or toxic springs.

I suspect biological life has conquered space already.   Life on earth may have originated by colonization from life that originated elsewhere.  There is no evidence for this.   But, given how successful life has been in replicating its will for conquering the entire planet, it is not hard to imagine that this will could predate life on earth.

I imagine biological life, in some form, may exist on icy bodies of asteroids and comets.   Within our solar system, the bulk of the biosphere may actually be on the icy bodies of the Kuiper belt or the Oort cloud.   There are abundant objects in this region although they are too small and dim for use to see.  Life could exist within them in microscopic niches where thin layers of liquid water may exist.

The idea of the universe being taken over by intelligent machines imagines a migration of machines from a earth-like planet that can support complex intelligent biological beings to create the machines.    In order for machines to exist, there had to some biological intelligence in the past to create them.

I think it is unlikely that the biological will to reproduce within the intelligent biological beings will be satisfied by sending out mechanical robots.  Instead these beings will send themselves out beyond their home planets just as their ancestors sent the first adventurers onto land from an overpopulated pool of water.   To conquer space, they will use machines to assist in that adventure.   The biological intelligent beings will colonize the icy bodies of the outer solar system.  They will mine the bodies for ores and for fusion fuels.   As their extra-terrestrial populations grow, they will expand to increasingly more distant objects.   Eventually they will expand to interstellar space until they are captured into an comparable distant object cloud of another star system and from there migrate closer to that star.

The dominant life form in the universe is more likely to be biological than machines.   The bulk of that life remains undiscovered at the edges of planetary systems, content to stay there indefinitely expanding from one object to the next.   Occasionally they will encounter a star system with a habitable planet.   Eventually they will reach that planet and introduce life to that planet if the planet is previously sterile.

So far in human history, we have a sent out artificial robots into deep space, landing on Mars, Venus, Titan, and plummeting into other worlds.  These machines lacked any capacity to reproduce.   However, with near certainty each has extremophile biological stowaways.  Those stowaways will reproduce as soon as they can find a chance.   Biology will always outnumber machines.

Advertisements

2 thoughts on “Superintelligent Robots lack biological will to live and reproduce

  1. Pingback: Dedomenocracy lessens the authority of expertise | kenneumeister

  2. Pingback: Dedomenocracy’s nemesis: the innovative criminal | kenneumeister

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s