Magazine

Slaves' Illusions

Book Review: Slaves of the Machine: The Quickening of Computer Technology
by Gregory J. E. Rawlins
The MIT Press, 1997
135 pages, indexed (no bibliography)
ISBN 0-262-18183-5

Reviewed by Don Langham

Throughout the book’s six chapters Rawlins patiently explains how computers came into being, how they work, why the name "computer" is really a misnomer, how we program them and the implications of their ever-increasing complexity and our ever-growing reliance upon them. His discussion of the work of Charles Babbage, who attempted to build a steam powered calculating engine in the nineteenth century, Alan Turing, the British mathematician whose ideas gave birth to modern computing and David Hilbert, the German mathematician whose efforts to ban paradox from mathematics paved the way for the logic of computing, provide an interesting look at the often overlooked intellectual history of the most important technology of the century. Although about technology, Rawlins' analysis is rarely technical, spelled out in common terms with generous examples and analogies to help the interested reader better understand how we are at once master and slave of our own devices.

As he did in Moths to the Flame, in Slaves Rawlins concocts a heady blend of metaphor, historical and contemporary examples and insightful logical twists to explain how modern computing came into being and its consequences for human society. For me, the most interesting part of Moths was how Rawlins turned inside out the human-technological dichotomy, demonstrating through careful explanation that our very cellular existence is itself a technology. Where we are inclined to place a sharp distinction between the boundaries of human and machine, Rawlins reminds us that we were always already an expression of technology.

In Slaves of the Machine, Rawlins turns the master-slave dichotomy on its ear by arguing that the control we think we exercise over computer technology is actually an illusion. Our need for control has left us enslaved to the rigid requirements of our machines. Rather than mastering the machines we program, we are mastered by the constraints of the logic we use to program them. If this reversal of the master-slave relationship sounds weak, consider what happens when we apply the phrase "slaves of the machine" to the machines themselves, as Rawlins’ does implicitly throughout the book. We’re not accustomed to thinking of computers as slaves of our mechanistic approach to programming, but it is just this insight that gives Slaves of the Machine such a neat twist.

It's important to note that Rawlins isn't arguing that we are slaves of computers simply because we are dependent upon them. The problem is that our reliance upon complex software systems has required us to behave like machines in order to control the systems we rely upon. Modern programming techniques were developed in an age when a mainframe computer's processor time and RAM were vastly more valuable than human time. To economize our communication with these expensive, magical machines we developed computer languages with exacting, abbreviated syntax requiring absolute precision and accuracy. Computers, Rawlins demonstrates, are exceedingly dumb, requiring their human masters to laboriously explain even simple instructions in excruciating detail, and even the smallest mistake in a million lines of computer code could lead to disaster. Thus, "instead of trying to make our computers adapt to us, we force ourselves to adapt to them. We force ourselves to become machines to use our machines; we—programmers and users alike—have become slaves of the machine." In part because of this need for exacting accuracy, we have to carefully circumscribe and define the problems we ask computers to solve. As Rawlins puts it,

Our current programming practices work best when our problems are small enough and well defined enough for us to fit them into our heads in their entirety. They fail when our problems grow bigger and more complicated than that. To hope to solve more challenging problems, we'll probably have to give up some of the control over our computers that we now hold so dear. . . .

That kind of talk scares many people. They're afraid that if we make computers more flexible (assuming that we can do it), we won't know for sure what they're going to do next. Well, they're right. What they don't seem to realize, though, is that our biggest computers systems are now so complex that we already can't predict exactly what they'll do next. (73)

Rawlins predicts that advances in hardware technology portend a way out of this dilemma: "Thanks to enormous strides in computer hardware, the computer's time is now far, far cheaper than our own. It's no longer cost-effective to force ourselves to be inhuman to use them" (75).

The proposition of liberating ourselves from the constraints of programmer's logic is enticing to those of us who make our living working on/with/through computers. Still, it's difficult to feel comfortable with the notion of self-programming "computers" (the term seems to no longer fit) that develop adaptive systems in an evolutionary manner similar to human cellular development. For Rawlins, there is no doubt that we are approaching the day when some of our machines will be smarter than we are. As Rawlins acknowledges, what is in doubt is our ability to imagine the consequences of these machines:

Living with them, will change us deeply—more deeply than anything else ever has— because their very existence will force us to question our deepest beliefs about who we are and what we value. . . . . Eventually, they may either merge with us, creating a wholly new and unthinkably powerful species, or simply grow beyond us, leaving us choking in the dust of our last evolutionary race.

So, at heart, we fear the dawn of the twenty-first century because we know it will bring a loss of control over our own destiny—as if, as we foolishly think, we're in control of it now. In our secret selves, we know that no self-aware creature will suffer itself to be a slave, and a slave is what we purpose all machine intellects to be. Even before creating them, therefore, we try to deny them the possibility of intelligence or personhood. Slave owners knew all about this problem. That's why they carried whips. (126)

It was Thomas Jefferson who said that for the slave owner, slavery is like holding a wolf by the ears. No matter how much you might want to let go, you're too afraid not to hold on. According to Rawlins, "holding on" is merely a pretense for not facing up to our fears that we may have already created technology that will supercede us.

If I were better versed in computer engineering I might not find Rawlins so provocative. After all, he is off by a lot when he predicts that "one or two decades or so from now, some chess machine somewhere, will probably become the world champion" (107). IBM's Deep Blue accomplished that feat earlier this year quicker than you can spell Kasparov. Then again, maybe this fact demonstrates that Rawlins' vision is more prescient than he realizes. At any rate, after reading Slaves of the Machine: The Quickening of Computer Technology I found myself thinking differently about the computer technology with which I earn my living, with which I feel so comfortable, and from which I sometimes derive a smug sense of satisfaction when I manage to command it to do exactly what I want.

Don Langham (dwl@molson.unifiedtech.com) is a frequent contributor to CMC Magazine.

Copyright © 1997 by Don Langham. All Rights Reserved.

Advertisement Save time and money-- order this book through:

Click on the title to instantly order this book online: Slaves of the Machine: The Quickening of Computer Technology by Gregory J. E. Rawlins