Sunday, April 2, 2017

Talking To Machines As If They Were People

            One of the cores ideas consistently discussed in this column has been mirroring.  A baby reaches out to his mother to satisfy his needs and his mother’s response to him is based on how she perceives him and what she feels for him.  The baby picks up on his mother’s attitudes as the mother mirror’s her child by the nature of her response to him.  Primarily positive mirroring leads to the baby’s development of a positive attitude towards himself.  Primarily negative mirroring leads to a negative attitude towards himself.  Such mirroring actually occurs, on some level, throughout a person’s youth and even later on, not only with parents but with other authority figures and even contemporaries, and has an enormous influence on how he evolves.

            In my own ruminations, I have concluded that a kind of mirroring occurs with the interaction between humans and another class of complex behavioral entity: namely modern complex machines.  Modern machines don’t have a coherent sense of self or a coherent consciousness, but they do have complex defined discrete responses to human attempts to control them and operate them.  To the extent that machines operate primarily with defined discrete processes, they operate best with instigation from defined discrete actions on the part of humans.  To the extent that humans get sloppy with their actions and start interacting with machines through more flowing, blendable continual behavior, through accidentally hitting the wrong keys on a computer or pressing the wrong buttons or hitting things too hard or for too long, they get poor responses from the machines.  This then reflects fully on the humans’ capacity to properly control and operate the machines and, by extension, on the humans’ effectiveness and potency  in the modern world.  Machines are not built to respond properly with the imprecision evinced in improperly targeted or managed flowing blendable continual actions on the part of humans.  To get optimum responses from a machine, humans receive the implicit mirroring that comes from poor performances as a result of sloppy operating of the machine and then start molding themselves after the machine in order to demonstrate the behavior that gets the machine to operate properly.

            It is this combination of mirroring and modeling involving modern machines that is leading to the gradual robotizing of humans in modern technological society.  But now comes voice control as the vehicle by which more and more modern machines will be controlled and operated.  Talking to modern machines could become as commonplace as talking to other humans and to domesticated animals.  Talking to machines to do things will replace using touch to manoeuver machines in order to complete our tasks.

            One more category of nuisance taken away from the human condition.  Right?  Voice control is one more step in the modern journey to remove friction from our lives.  Get rid of friction and you can get rid of stress.  Get rid of stress and you can be a happy person.  At least, one can say, that this is what is implied in our modern philosophy of life.

            But actually people need a certain amount of friction in the real world in order to feel fully alive.  Get rid of too much organic friction, and a person sinks into a sensory vacuum, an experiential void.  Because touch involves direct sensory rubbing or grating with tactile phenomena in the external world, there is a lot more immediate organic friction involved in doing things with one’s hands than with one’s voice.  Speech is a much more mediated interaction with phenomena in the external world than manual manipulation.  Speech was not created for direct manipulation of the external world, but rather for communication with other humans and with domesticated animals.  One might say that speech traditionally has acted sometimes as a complement to manual manipulation as when riders of animals like horses or of chariots, carriages or stagecoaches driven by animals have talked to their animals in conjunction with their actions to either spur them on or to make them stop.  Speech has also been used by slaveholders to order around their slaves, by generals to order around their soldiers, and by bosses to order around their workers.  But this is not the same as what is happening today where speech is replacing manual manipulation not only for control of other organisms, but for all sorts of direct control of different tasks in the external world.

            So what will a transition from manual manipulation to voice control do to how people experience not only the technology and other physical phenomena in their world but also themselves?

            The use of voice control means that the form of sensory engagement that was once used for communicating with humans and animals in order to fine-tune relationships or to coordinate group projects that involved some manual manipulation is now being used for the direct manipulation by individual people of the things around them.  But because of the sophistication of many of these things, namely complex machines like computers, smartphones and robots, some of the vocal engagements with these complex things involve interactions that approximate dialogues with humans.  Obviously, the technology is still evolving and efforts are being made to increase both voice recognition and vocabulary on the part of these machines.  To the extent that these dialogues between humans and machines start to increase, the line that separates other humans from machines in the mind of a communicating human begins to blur.  Machines are increasingly being endowed with many of the qualities of humans (and humans are being endowed with many of the qualities of machines).  We increasingly are acting with machines as if they had a coherent sense of self and a coherent consciousness.

            And to the extent that we implicitly endow machines with human attributes because of their response to human vocalizing or their capacity to vocalize themselves, we become susceptible to mirroring from the machine (the way the machine appears to respond to our strengths and limitations) as well as modeling on our part with respect to an implicitly idealized crisp well-defined discrete relatively smooth frictionless performance on the part of the machine.  When we vocalize to a machine, we are treating it like a human being.

            By the same token, we start to view people from the perspective of their functional instrumental machine-like qualities.  How good are they at performing the narrow functional tasks that are required of them?  Just pay people for their services at a given time and don’t worry about their needs as whole human beings.  In the modern U.S., companies try to get workers on a part-time basis, as independent contractors for particular jobs or projects in order to avoid having to pay for benefits like health care, time off for child birth and pensions.  During those times when a worker is not actually functioning for his company or companies, he is on his own.  For young men and increasingly young women, there is a growing tendency to focus on sexual performance in their partners and to be less interested in emotional bonds that can create grounded relationships.  As a matter of fact, increasingly all relationships become contingent on performance rather than providing the foundation for a more unconditional connection.


            These are some of the developing deleterious effects that are at least partly due to relating to modern complex machines through voice as if the machines were human.  I have talked about the blurring of the lines between humans and machines as manifested in androids and cyborgs.  But voice control represents a more subtle functional blurring of lines between humans and machines.  It is nevertheless a blurring that is every bit as pernicious and one that threatens the very essence of our human identity.

(c) 2017 Laurence Mesirow

No comments:

Post a Comment