"The wish to impose order upon confusion, to bring harmony out of dissonance and unity out of multiplicity is a kind of intellectual instinct, a primary and fundamental urge of the mind. Within the realms of science, art and philosophy the workings of what I may call this "Will to Order" are mainly beneficent. True, the Will to Order has produced many premature syntheses based upon insufficient evidence, many absurd systems of metaphysics and theology, much pedantic mistaking of notions for realities, of symbols and abstractions for the data of immediate experience. But these errors, however regrettable, do not do much harm, at any rate directly—though it sometimes happens that a bad philosophical system may do harm indirectly, by being used as a justification for senseless and inhuman actions. It is in the social sphere, in the realm of politics and economics, that the Will to Order becomes really dangerous.
Here the theoretical reduction of unmanageable multiplicity to comprehensible unity becomes the practical reduction of human diversity to subhuman uniformity, of freedom to servitude. In politics the equivalent of a fully developed scientific theory or philosophical system is a totalitarian dictatorship. In economics, the equivalent of a beautifully composed work of art is the smoothly running factory in which the workers are perfectly adjusted to the machines. The Will to Order can make tyrants out of those who merely aspire to clear up a mess. The beauty of tidiness is used as a justification for despotism.
Organization is indispensable; for liberty arises and has meaning only within a self-regulating community of freely co-operating individuals. But, though indispensable, organization can also be fatal. Too much organization transforms men and women into automata, suffocates the creative spirit and abolishes the very possibility of freedom. As usual, the only safe course is in the middle, between the extremes of laissez-faire at one end of the scale and of total control at the other.
During the past century the successive advances in technology have been accompanied by corresponding advances in organization. Complicated machinery has had to be matched by complicated social arrangements, designed to work as smoothly and efficiently as the new instruments of production. In order to fit into these organizations, individuals have had to deindividualize themselves, have had to deny their native diversity and conform to a standard pattern, have had to do their best to become automata."
As the previous quote makes clear, Huxley, in 1958, was concerned that the world was becoming too organized. That is, the behavior of groups of humans was becoming more and more organized to achieve the interests of some smaller group of humans, or potentially of the emergent interests of the organizations themselves. Less pronounced in the chosen quote above, but key to Huxley's conceptualization of the Brave New World of his dystopia is that the most effective way of securing this organized behavior is through positive rewards. It is through the systematic offering of these positive rewards that, in the long run, large organizations (say Big Government and Big Business, the forms or organization Huxley points to), systematically manipulate human behavior by offering sets of positive rewards that are in the interest of some minority controlling group or of the emergent goals of the organization itself.
This insight is, of course, not unique to Huxley. In fact, it is a key component of organization theory itself. Herbert Simon, one of the early modern explorers of both organization theory and artificial intelligence, described organizations as offering inducements (some set of positive rewards) to individuals to induce them to participate in the mission of the organization, whether that be the profit seeking motive of a company or the mission of a government agency.
These observations point to the fact that throughout history human organizations have required the humans within them to sacrifice some of their own goals, wishes, and desires, to receive some positive reward from some form of compensation. The level and type of sacrifice has changed over time. For example, at points in history, the CHURCH or the EMPIRE were large organizations that demanded complete or near complete sacrifice on behalf of one's own goals and instead required the individual to fully adopt those of the organization for the supposed positive rewards of eternal life and eternal glory.
To be fair, organizations themselves often do indeed provide very meaningful positive benefits to the individual in the way of community, social life, and specialization in the face of complexity. In fact, modern life, and much of pre-modern life could not be maintained without the organizations of families, churches, governments, businesses, empires, etc. Organizations offer individual humans positive rewards for their participation, and the organizations themselves offer these positive rewards so that the individual humans will work towards the goals of the organization. As Huxley reminded us above, organization is indeed indispensable to freedom, but taken to extremes it kills individual autonomy and thus freedom.
Now, let me use some examples where individuals behave in bad ways to serve the ends of a larger organizations for some promised positive reward, obvious cases include:
-Honor killings of family members (to ensure the family's honor)
-Extreme political violence (to serve one's Political Party and receive glory)
-Corruption on behalf of corporate executives (to make money and have power)
These examples (I hope) make self-evident the logic of organizations offering positive rewards to individuals to achieve the organization's goals, while inducing the human to behave in ways that we generally agree are unethical and generally harmful to their fellow humans, and certainly denigrate the humanity of the individual engaged in these activities.
It is here that the language of alignment is appropriate to adopt. Organizations seek to align the behavior of individual humans with the goals of the organization. In this way, the organization's goals are put ahead of the needs and flourishing of the individuals within those organizations. That is individual flourishing is sacrificed to the alignment or adaptation or conformity that the organization seeks for its own purposes.
Aligning Organizations and Humans with Algorithms
Let's examine a large modern, multinational organization, that seemingly has come to influence the entire digitally connected world: Facebook. Here is an organization that grew exponentially and created an input supply chain of several billion humans for which it provided the positive reward of "social connection" that is organized by the platform's advertising algorithm. Here, the goal of the organization is mass, individualized advertising through one of the most powerful algorithms in the world. The humans that flow through this organization are both employees and the input supply chain of human behavior. In this case the advertisers purchase individualized attention space based upon prediction from the aggregate supply chain that is tailored to individual profile types by the desires of the advertisers. The organization structure, the controlling group, the positive rewards, they are tangled in a different sort of web than the one upon which the traditional CHURCH or EMPIRE were tangled.
In this case it is a mechanical, digital, algorithm that is shaping both the behavior of the organization and of individual users. In this way the algorithm itself can be thought of as the controlling group, controlling the behavior of its users. The controlling group (the algorithm), offers positive rewards to its customers (the advertisers), for shaping the behavior of its users (individuals with profiles on the platform). For a detailed account of this, read Shoshana Zuboff's The Age of Surveillance Capitalism.
The important distinction here is from the controlling group being some minority group of humans shaping and directing the behaviors of the individuals within an organization, to one where an algorithm has assumed much of this controlling power reshapes the control mechanisms and the power of the will to order once it is mechanized and directly tailored to the weaknesses of its users. In this way, the consequences of over-organization can be seen not only by "the organization man" who is an employee of an organization, but also to all the users of that platform. If the users can be meaningfully organized into group A and group B, then ads can be more intelligently crafted to BOTH group A and group B, and this is much more efficient for the advertisers to deliver effective ads. The blue group will be more clearly shaped and defined in its preferences, as will the red group, and as a result individual preferences become aligned with those of the organization and its controlling entity, the algorithm.
It seems that this shift, is accompanied by a shift from "the organization man" as the product of the organization for which he toils, to a spreading of the organization man ethos throughout society. And this ethos of sacrificing ones own individuality to the ends of the organization for positive rewards has a direct analogy in shaping the behaviors and reducing the individuality of those who meaningfully engage with the algorithm controlled organization.
The Organization Man
In the mid-20th century, while Huxley was making the argument that the world was beginning to suffer from over-organization, another scholar of the time, William Whyte, had documented what this meant for the men (and at that time the organizations under study were dominated mostly by men) inhabiting the large organizations that had developed throughout much of the world, but particularly so in the United States.
Whyte studied the ethos of the individuals (again in this case mostly men) of his time that were aligning themselves to modern organization life. In his treatise, The Organization Man, Whyte notices a change in the culture of organization life in the post-WWII USA. One particular point he highlights is that while this tension of organizational inducements for the behavior of individuals is certainly not a new phenomena, he is taken aback by what he sees as a lack of recognition of this tension. For Whyte, it is given that there are important tensions for what is good for the individual and what is good for the organization, but his concern is that for many individuals, in adapting or aligning themselves with an organization, there often seemed to be a lack of recognition of this tensions existence. That is rather than individuals wrestling with the complex tradeoffs across the positive rewards offered by an organization, their ethos became one of alignment with the organization for the sake of alignment itself. That is, the organization, rather than the individual, defines what is and what should be good for the individual. It is the idea that organization is good in and of itself, that the good of the group is inherently more important than the good of the individual, and that individuals can only achieve the good through submitting to the group and to the organization.
Whyte even imagined a time in which this would infiltrate the work of scientists, with this prophetic statement:
"Suppose for the moment that you were given this mental exercise: without knowing anything about how scientists work today, you were to imagine what would happen if the Social Ethic were applied to science as it has been in the rest of organization life. The chances are that you would imagine, among other things, that: (1) scientists would now concentrate on the practical application of previously discovered ideas rather than the discovery of new ones; (2) they would rarely work by themselves but rather as units of scientific cells; (3) organization loyalty, getting along with people, etc. would be considered just as important as thinking; (4) well rounded team players would be more valuable than brilliant men, and a very brilliant man would probably be disruptive. Lastly and most important, these things would be so because people believe this is the way it should be."
And, one need not look very far into the lives of modern scientists working within modern universities to see how far this ethos has infiltrated even this group, once renowned for its focus on the individual and the power of individual work and individual discovery, to see the current dominating structure of conformity to practical applications, working as units, organization loyalty, and being a well rounded team player.
Over-Organization & The Algorithmically Aligned Human
In the technical AI alignment and AI safety community, there is concern about how to successfully align machine intelligence with that of humanity and human values. Much of this work is concerned with the challenge of controlling increasing intelligent machines to behave and act as we would prefer they did. The concern is that once a powerful enough machine is built that is unaligned with the interests of humans, that this could lead to catastrophe for humanity, because humanity would lose control of its future and effectively be controlled by the machine intelligence.
However, less is made of the fact that right now, with the widespread deployment of organization controlling algorithms, that we are already in the midst of a process that is designed to align humans with the goals of the algorithm. That is, there are social and economic pressures throughout the world that are pushing alignment in the wrong direction. We think of humans as the "principals" and the algorithms as the "agents" that we are attempting to align, but, in fact, this is the wrong analogy to understand the current way in which human are being aligned with the goals of algorithms. In fact, we have a situation where the algorithms themselves have often become the "principals" and they are finding ever more powerful ways to align the human "agents."
And this is where over-organization leads us to an examination of over-digitalization. For much of history, individuals have been set against organizations in their battle for individuality and for freedom. This battle became increasingly difficult in the age of Big Business and Big Government, but at the top of these organizations there were at least other humans making decisions that responded to a variety of other human pressures, but we seem to have marched or maybe sleepwalked into a circumstance where algorithms have often become the effective controlling group and it is these algorithm pressures and organization pressures that the individual, that wishes to remain an individual, must set themselves against. And, as it turns out, the toxic mixture of controlling algorithms and controlling organizations may snuff out the individual impulse to individuality and to freedom, killing with it the human spirit.
The Will to Order
"The wish to impose order upon confusion, to bring harmony out of dissonance and unity out of multiplicity is a kind of intellectual instinct, a primary and fundamental urge of the mind. Within the realms of science, art and philosophy the workings of what I may call this "Will to Order" are mainly beneficent. True, the Will to Order has produced many premature syntheses based upon insufficient evidence, many absurd systems of metaphysics and theology, much pedantic mistaking of notions for realities, of symbols and abstractions for the data of immediate experience. But these errors, however regrettable, do not do much harm, at any rate directly—though it sometimes happens that a bad philosophical system may do harm indirectly, by being used as a justification for senseless and inhuman actions. It is in the social sphere, in the realm of politics and economics, that the Will to Order becomes really dangerous.
Here the theoretical reduction of unmanageable multiplicity to comprehensible unity becomes the practical reduction of human diversity to subhuman uniformity, of freedom to servitude. In politics the equivalent of a fully developed scientific theory or philosophical system is a totalitarian dictatorship. In economics, the equivalent of a beautifully composed work of art is the smoothly running factory in which the workers are perfectly adjusted to the machines. The Will to Order can make tyrants out of those who merely aspire to clear up a mess. The beauty of tidiness is used as a justification for despotism.
Organization is indispensable; for liberty arises and has meaning only within a self-regulating community of freely co-operating individuals. But, though indispensable, organization can also be fatal. Too much organization transforms men and women into automata, suffocates the creative spirit and abolishes the very possibility of freedom. As usual, the only safe course is in the middle, between the extremes of laissez-faire at one end of the scale and of total control at the other.
During the past century the successive advances in technology have been accompanied by corresponding advances in organization. Complicated machinery has had to be matched by complicated social arrangements, designed to work as smoothly and efficiently as the new instruments of production. In order to fit into these organizations, individuals have had to deindividualize themselves, have had to deny their native diversity and conform to a standard pattern, have had to do their best to become automata."
-Aldous Huxley, Brave New World Revisited
Aligning Humans With Organizations
As the previous quote makes clear, Huxley, in 1958, was concerned that the world was becoming too organized. That is, the behavior of groups of humans was becoming more and more organized to achieve the interests of some smaller group of humans, or potentially of the emergent interests of the organizations themselves. Less pronounced in the chosen quote above, but key to Huxley's conceptualization of the Brave New World of his dystopia is that the most effective way of securing this organized behavior is through positive rewards. It is through the systematic offering of these positive rewards that, in the long run, large organizations (say Big Government and Big Business, the forms or organization Huxley points to), systematically manipulate human behavior by offering sets of positive rewards that are in the interest of some minority controlling group or of the emergent goals of the organization itself.
This insight is, of course, not unique to Huxley. In fact, it is a key component of organization theory itself. Herbert Simon, one of the early modern explorers of both organization theory and artificial intelligence, described organizations as offering inducements (some set of positive rewards) to individuals to induce them to participate in the mission of the organization, whether that be the profit seeking motive of a company or the mission of a government agency.
These observations point to the fact that throughout history human organizations have required the humans within them to sacrifice some of their own goals, wishes, and desires, to receive some positive reward from some form of compensation. The level and type of sacrifice has changed over time. For example, at points in history, the CHURCH or the EMPIRE were large organizations that demanded complete or near complete sacrifice on behalf of one's own goals and instead required the individual to fully adopt those of the organization for the supposed positive rewards of eternal life and eternal glory.
To be fair, organizations themselves often do indeed provide very meaningful positive benefits to the individual in the way of community, social life, and specialization in the face of complexity. In fact, modern life, and much of pre-modern life could not be maintained without the organizations of families, churches, governments, businesses, empires, etc. Organizations offer individual humans positive rewards for their participation, and the organizations themselves offer these positive rewards so that the individual humans will work towards the goals of the organization. As Huxley reminded us above, organization is indeed indispensable to freedom, but taken to extremes it kills individual autonomy and thus freedom.
Now, let me use some examples where individuals behave in bad ways to serve the ends of a larger organizations for some promised positive reward, obvious cases include:
-Honor killings of family members (to ensure the family's honor)
-Extreme political violence (to serve one's Political Party and receive glory)
-Corruption on behalf of corporate executives (to make money and have power)
These examples (I hope) make self-evident the logic of organizations offering positive rewards to individuals to achieve the organization's goals, while inducing the human to behave in ways that we generally agree are unethical and generally harmful to their fellow humans, and certainly denigrate the humanity of the individual engaged in these activities.
It is here that the language of alignment is appropriate to adopt. Organizations seek to align the behavior of individual humans with the goals of the organization. In this way, the organization's goals are put ahead of the needs and flourishing of the individuals within those organizations. That is individual flourishing is sacrificed to the alignment or adaptation or conformity that the organization seeks for its own purposes.
Aligning Organizations and Humans with Algorithms
Let's examine a large modern, multinational organization, that seemingly has come to influence the entire digitally connected world: Facebook. Here is an organization that grew exponentially and created an input supply chain of several billion humans for which it provided the positive reward of "social connection" that is organized by the platform's advertising algorithm. Here, the goal of the organization is mass, individualized advertising through one of the most powerful algorithms in the world. The humans that flow through this organization are both employees and the input supply chain of human behavior. In this case the advertisers purchase individualized attention space based upon prediction from the aggregate supply chain that is tailored to individual profile types by the desires of the advertisers. The organization structure, the controlling group, the positive rewards, they are tangled in a different sort of web than the one upon which the traditional CHURCH or EMPIRE were tangled.
In this case it is a mechanical, digital, algorithm that is shaping both the behavior of the organization and of individual users. In this way the algorithm itself can be thought of as the controlling group, controlling the behavior of its users. The controlling group (the algorithm), offers positive rewards to its customers (the advertisers), for shaping the behavior of its users (individuals with profiles on the platform). For a detailed account of this, read Shoshana Zuboff's The Age of Surveillance Capitalism.
The important distinction here is from the controlling group being some minority group of humans shaping and directing the behaviors of the individuals within an organization, to one where an algorithm has assumed much of this controlling power reshapes the control mechanisms and the power of the will to order once it is mechanized and directly tailored to the weaknesses of its users. In this way, the consequences of over-organization can be seen not only by "the organization man" who is an employee of an organization, but also to all the users of that platform. If the users can be meaningfully organized into group A and group B, then ads can be more intelligently crafted to BOTH group A and group B, and this is much more efficient for the advertisers to deliver effective ads. The blue group will be more clearly shaped and defined in its preferences, as will the red group, and as a result individual preferences become aligned with those of the organization and its controlling entity, the algorithm.
It seems that this shift, is accompanied by a shift from "the organization man" as the product of the organization for which he toils, to a spreading of the organization man ethos throughout society. And this ethos of sacrificing ones own individuality to the ends of the organization for positive rewards has a direct analogy in shaping the behaviors and reducing the individuality of those who meaningfully engage with the algorithm controlled organization.
The Organization Man
In the mid-20th century, while Huxley was making the argument that the world was beginning to suffer from over-organization, another scholar of the time, William Whyte, had documented what this meant for the men (and at that time the organizations under study were dominated mostly by men) inhabiting the large organizations that had developed throughout much of the world, but particularly so in the United States.
Whyte studied the ethos of the individuals (again in this case mostly men) of his time that were aligning themselves to modern organization life. In his treatise, The Organization Man, Whyte notices a change in the culture of organization life in the post-WWII USA. One particular point he highlights is that while this tension of organizational inducements for the behavior of individuals is certainly not a new phenomena, he is taken aback by what he sees as a lack of recognition of this tension. For Whyte, it is given that there are important tensions for what is good for the individual and what is good for the organization, but his concern is that for many individuals, in adapting or aligning themselves with an organization, there often seemed to be a lack of recognition of this tensions existence. That is rather than individuals wrestling with the complex tradeoffs across the positive rewards offered by an organization, their ethos became one of alignment with the organization for the sake of alignment itself. That is, the organization, rather than the individual, defines what is and what should be good for the individual. It is the idea that organization is good in and of itself, that the good of the group is inherently more important than the good of the individual, and that individuals can only achieve the good through submitting to the group and to the organization.
Whyte even imagined a time in which this would infiltrate the work of scientists, with this prophetic statement:
"Suppose for the moment that you were given this mental exercise: without knowing anything about how scientists work today, you were to imagine what would happen if the Social Ethic were applied to science as it has been in the rest of organization life. The chances are that you would imagine, among other things, that: (1) scientists would now concentrate on the practical application of previously discovered ideas rather than the discovery of new ones; (2) they would rarely work by themselves but rather as units of scientific cells; (3) organization loyalty, getting along with people, etc. would be considered just as important as thinking; (4) well rounded team players would be more valuable than brilliant men, and a very brilliant man would probably be disruptive. Lastly and most important, these things would be so because people believe this is the way it should be."
William Whyte, The Organization Man, 1956
And, one need not look very far into the lives of modern scientists working within modern universities to see how far this ethos has infiltrated even this group, once renowned for its focus on the individual and the power of individual work and individual discovery, to see the current dominating structure of conformity to practical applications, working as units, organization loyalty, and being a well rounded team player.
Over-Organization & The Algorithmically Aligned Human
In the technical AI alignment and AI safety community, there is concern about how to successfully align machine intelligence with that of humanity and human values. Much of this work is concerned with the challenge of controlling increasing intelligent machines to behave and act as we would prefer they did. The concern is that once a powerful enough machine is built that is unaligned with the interests of humans, that this could lead to catastrophe for humanity, because humanity would lose control of its future and effectively be controlled by the machine intelligence.
However, less is made of the fact that right now, with the widespread deployment of organization controlling algorithms, that we are already in the midst of a process that is designed to align humans with the goals of the algorithm. That is, there are social and economic pressures throughout the world that are pushing alignment in the wrong direction. We think of humans as the "principals" and the algorithms as the "agents" that we are attempting to align, but, in fact, this is the wrong analogy to understand the current way in which human are being aligned with the goals of algorithms. In fact, we have a situation where the algorithms themselves have often become the "principals" and they are finding ever more powerful ways to align the human "agents."
And this is where over-organization leads us to an examination of over-digitalization. For much of history, individuals have been set against organizations in their battle for individuality and for freedom. This battle became increasingly difficult in the age of Big Business and Big Government, but at the top of these organizations there were at least other humans making decisions that responded to a variety of other human pressures, but we seem to have marched or maybe sleepwalked into a circumstance where algorithms have often become the effective controlling group and it is these algorithm pressures and organization pressures that the individual, that wishes to remain an individual, must set themselves against. And, as it turns out, the toxic mixture of controlling algorithms and controlling organizations may snuff out the individual impulse to individuality and to freedom, killing with it the human spirit.
**********
This has been Chapter 5 in the From Our Brave New Word to Analogia Sequence.
Chapter 1, Chapter 2, Chapter 3, and Chapter 4 are each available as well.
Chapter 6 is also now available.