A History and Future of the Rise of the Robots

We’re heading towards the final fulfillment of the Industrial Revolution.

We take time for granted. Not in the “live like it was your last day on earth” kind of way (though likely in that way too). If you’re wearing a watch, look at it. Does it have a second hand? Why? Just in case you suddenly need to stand in as time keeper for an Olympic sprint? Does it have a minute hand? Almost certainly. Time, seen as such, is a relatively new and novel contribution to humanity. It wasn’t a technical issue. The ancients were no dummies when it came to coming up with solutions to their pressing problems. The Chinese and the Vikings figured out precise compases over 1000 years ago. Ancient Roman concrete was vastly superior to our own, and in China they had developed a surprisingly accurate and effective seismoscope to detect distant earthquakes by AD 132! In fact, even precise mechanical clocks had been invented by 725 AD, yet somehow didn’t catch on for another thousand years. Why?
We take time for granted as seconds ticking away at 60 clicks a minute, 60 minutes per hour, 24 hours a day and so on. This hasn’t always been the case. Even when the technology was available for such accurate clocks, why didn’t it catch on for so long, and what changed? E. P. Thompson, the renowned historian who changed the field of social history with his seminal book “The Making of the English Working Class,” also explored how our view of time changed alongside our approach to work. The clock may seem like a rather mundane object of focus, especially when so many of us now keep ours in our purses and pockets. This isn’t just talking about clocks and watches. I’m not even primarily talking about clocks and watches. Much has been said about automation and the replacement of the human workforce for decades now. Lately it’s reached a new height. Autonomous driving is on the horizon. Donald Trump surprised the world, most critically via the votes of the rust belt “working poor” and underemployed, many of whom lost their jobs directly or indirectly to robots.
Our concept of time is inseparably linked with our careers as components, cogs, and tools in the larger systems which drive our economies, provide for our needs, and fulfill our wants. Since the industrial revolution, It’s been just long enough for us take our jobs, careers, and work for granted, and thus fail to see, baring any unforeseen catastrophes, the unsustainability of such an arrangement. Put simply, the last few hundred years have been an anomaly in the history of working humans. It’s vital to explore the past, present, and future of work shifting from humans to machines, how it has impacted our view of time, and take a glimpse into our future.

Before the Machines

Thompson describes a world, before the industrial revolution, where work primarily functioned under what he refers to as “task-oriented” time. The task-oriented approach to time essentially sees time through the lens of the natural work-rhythms of the tasks the individual is attending to. Sheep must be herded, cows must be milked, shoes must be made, the hot coals must be tended and so forth. He describes there being very little demarcation between “work” and “life” at this time. Pause and think about that for a moment. Much ink has been spilt and money spent in the pressing quest to find the ideal “work/life balance,” or to keep work from “encroaching” on our personal lives. Thompson describes the social and the labor as “intermingled- the working-day lengthens or contracts according to the task- and there is no great sense of conflict between labor and “passing the time of day.”” He also points out that to those accustomed to labor by the clock (that’s us), this attitude “appears to be wasteful and lacking in energy.” It seems no great leap to suggest that in our current Western world this orientation to time is all but dead. How did we get from there to here?
I don’t want to romanticize the peasant life of the middle ages. It has been interesting for me, however, to realize how hard it is to even write about such an approach to time without seeming wistful as if longing for a “bygone” era. The reality is, our approach to time has changed. And it changed along with industrialization. Put another way, those of the pre-industrial “task-oriented” times, largely peasants or craftsmen, had direct ownership of their time. They were largely responsible for completing entire projects, such as making shoes, tending sheep, or building a wagon from scratch. They had the skills and the manual tools to undertake complete projects and arranged their day to day according to their need. The cobbler had complete control over when he worked, or didn’t, on the shoes. He may produce as many shoes as he is able in a day, or he may go an indefinite amount of time doing no work on producing shoes, depending on need and desire for goods, money, or shoes. This direct ownership comes because the cobbler had the knowledge and ability to create finished shoes from raw materials himself, and in part because he was the absolute master of his tools. He may, for example, use a blade to cut the leather, but the blade can do nothing of itself and so is simply an extension off the cobbler himself. While this seems mundane it’s an important differentiator to now. While he may never be wealthy, and within the scope of the larger market may or may not even be able to sustain a family with his cobbling, within the scope of his work he is the absolute master of his time. To better understand how machines have crept in, we’ll follow this cobbler through a remarkably long life from pre-industrialization to the largely post-human future workforce.

Introducing the Machines

As technological mechanization increased, the cobbler found himself in a different relationship to both his tools and his time. Rather than tools simply being an inert extension of his own self and expertise, they began to make demands on his time. We can, for example, imagine an automatic leather cutter recently acquired by the cobbler which allows him to simply put in a square piece of leather, wait a period, and then retrieves the newly shaped piece of leather. Previously this may have taken the cobbler 30 minutes to do. Now the machine can do it in 30 seconds. Furthermore, while his old manual blade could never force him to wait, the automatic cutter can. This surrender over control of his work time will almost certainly be seen by the cobbler as returning valuable time to him. He is now able to complete the total project faster, and still largely retains direct control of the entire process.
Alfred Chandler, a former professor of Business History at Harvard, helps provide some valuable insight into the next stage, and won the pulitzer prize in in doing it. Chandler describes how the new tools and machinery of the industrial revolution made production possible at vastly grander scales, and divided the work. Where once our cobbler was the master of the shoemaking process, the machines are now able to do too much work too quickly for the cobbler to effectively use them himself. What good is a machine that can cut 500 leather pieces a minute when the cobbler can only use three of them in an hour? In response to these new abilities, the complete shoemaking processes began to be segmented and specialized. The complete work of our cobbler, because of advances in manufacturing and shipping technology, has now been divided up across many different machines and workers housed at many different factories scattered all over. Specialization necessarily demands coordination, and that has dramatic implications for our cobblers time as we watch him move into the city and get a job at one of these factories.

Working With Machines

At this phase the humans and machines work together side by side as tools in a large and distributed production process of specialized tasks. Neither the machine nor the human directly owns their time, but must bow to the demands of coordination. A wide web of distributed sole-makers, leather cutters, sowers, designers, and assemblers are now coordinated to complete the same project previously completed by one cobbler. The new system, however, is now able to do it with dramatically higher consistency and output. For coordination to work, each part of the process must be timed to coincide with the other separate functions. Soles must be made, on time, to ship, on time, so that they may be combined with leather uppers, on time, at a different location, and so on. If the automatic cutting machine made minor constraints on when the cobbler couldn’t work, now the schedule constraints when he must. And so, as Thompson describes, the clock is given a minute hand, and then a second hand, is perfected and miniaturized, and then given by the companies to their employees for them to wear on their wrists. The human and the machine labor equally under their new master, the schedule.
It’s beyond typical in our era to experience time as “pressure,” and then to lament or brag about the “stress” of it all. A lot of this is the price of coordination, and the pressure is real. If any tool, human or machine fails to meet their scheduled output then the process is broken and that tool will be repaired or replaced. There is functionally very little difference at this point from the human and the tool. Put another way, humans are now simply the semi-autonomous tools of the larger production and distribution processes. Time is of existential importance to the tools in the project as any human or machine that continually fails to keeps its deadlines will almost certainly be replaced with one who will. And thus the work of mankind completes its shift from task orientation to the scheduled society.

The Present

This is the phase where most of us currently labor. We almost all strive as tools within a process that far exceeds our own work, and demands coordination with a multitude of other moving parts. I say a tool because a tool performs a task, but not a project. If a humanoid robot dressed just like our cobbler walked into his workshop and made shoes from scratch, just as our cobbler had done, he would not be a tool in the shoemaking process, he, as the cobbler before him, would be the shoemaking process. Humans now undertake work largely as tools within much larger projects/tasks, and are working side by side with machines, under the schedule, virtually everywhere. I’m not lamenting scheduling. Coordination can make wonderful things possible. It is, however, important to understand how the massively complex and intertwined systems of coordination which service the processes of production, distribution, and consumption have dramatically impacted our relationship to time, and also have implications for the future.
Central to the plight of the underemployed and working poor is the reality that there is nothing to suggest that this position is either stable or sustainable. We humans are largely only still involved in the process because we’re still the cheapest option for whatever task we’re doing. Cheaper because the technology is currently too expensive or non-existent, and cheaper because wages can always be lowered. As technology advances, however, humans are increasingly less effective and more expensive than good machines. This is true not just for those working at the ground floor, but also for the managers above them.

Moving Towards the Future

As the machine tools become increasingly automated, human labor is susceptible, and the position of the manager as tool is reinforced as they are as susceptible to replacement as all other human tools. Management, one of the largest professions in the Western world, becomes irrelevant through two separate but simultaneous processes. The first process occurs indirectly through attrition via a reduction of humans in the workplace for them to manage. The other process happens directly, through increasingly powerful and effective algorithmic or technological managers. Think of Uber. Unlike us, machines can always be described as “willing” to do what they are told, though historically not usually as able. There may be a broken part or a bug in the code, but when capable, they will always perform as asked. This is strikingly different from most management training today, which could be described largely as as “how to deal with the pesky problem of the disobedient, undertrained, and unmotivated human tools still necessary to your processes.”
Much of the initial replacement of human tools by machine tools, however, is largely not yet a replacement but a shift towards more service oriented work. The service sector in the United States accounts for over 78% of the total employment in the United States, and is growing rapidly while employment in production is not growing at all. To mix metaphors, we haven’t really been kicked out of the factory yet, (except in certain regions such as the rust belt), but have simply been moved down the conveyor line to the jobs that machines heretofore have not yet done very well. This, again, is not sustainable. The shift is primarily because of the difficulty of creating effective service-providing technology, and is temporary as machines can eventually be concocted to provide those services as well. For example, think of customer service call centers. Once you called and spoke to someone from your own country, then often to someone from another country, and now increasingly, you may ask your question in conversational language and have it answered by the speech recognizing robot on the other end, all without so much as a minute spent on hold.
As artificial intelligence and machine learning develops, particularly in their ability to understand and contribute in natural human conversation, humans will reach the end of their usefulness in an increasing number of industries systems entirely. Over 8 million people in the United States are employed in the trucking industry, for example. Otto, recently acquired by Uber, is already testing autonomous trucks on the roads. Tesla has announced plans to venture into the autonomous semi-truck business. Peloton Technologies has developed trucking software which allows these trucks to “link” up in lines on the highway saving tremendous amounts of time and gas. By nearly all accounts autonomous trucking is expected to be adopted much more quickly than consumer vehicles, which themselves appear only a few years from introduction.

The Future

Moving into the not so distant future, our cobbler lost his job in the factory about 50 years ago but luckily was able to get a job in a suburban mall selling shoes at Foot Keeper. Doing such a fine job on the floor he was promoted to supervisor and eventually store manager. Foot Keeper, however, is phasing out the store model. You peruse through various styles online and pick your favorite. You then throw down your phone and take 3D scans of your feet which are sent directly to the Foot Keeper factory filled with 3D printers which custom print the shoe, based on the unique shopper’s foot, in less than one minute. A machine then packages the shoes and loads them on a waiting Amazon Drone to be delivered to your doorstep this afternoon. (This scenario isn’t nearly as far away as it may seem. Look up Carbon3D). Alas, our cobbler has been completely cut out of the system. There may be other industries for him to jump into for a while as they mature into full automation, but for how long?
There is no reason to believe that human labor will somehow prevail as machine labor becomes more efficient, effective, and cheaper. It’s not the machines that are driving the take-over, it’s simply profit maximizing decisions by the company owners. Leaping all of the way into science fiction territory, in full automation the giant processes of segmented tasks can be subsumed under one great master machine. Think of 3D printers, even in their current infant state, and their remarkable ability to take an entire manufacturing process that used to be distributed between countless factories, and combine it into one small box. Such meta-systems are no longer an ecosystem of interdependent tasks, but are instead an aspen grove of various, seemingly separate, but internally singular manifestations of a sole working machine. In one final grand humanless resolution, the processes of production and distribution are once again subsumed by one machine process, scarcely different from the original cobbler in function, but worth millions of cobblers in output.
On the scale of only a few hundred years we are completing something of a circle. First, cobblers were the masters of the shoe making process, made shoes for themselves and their neighbors, and arranged their time accordingly. Then machines allowed unskilled, but punctual, factory workers to use machines to make shoes for everyone in mass. Eventually the machines got so good that they could largely make the shoes themselves and only a few humans are needed to stand by as technicians for when something goes wrong or the machines need to be recoded in order to learn new things. Even those technician positions are temporary. Industry is already rife with automated repair systems and machines. The May 2016 cover of Wired Magazine even announced the end of coding as machines become capable of coding and recoding themselves with ever more sophisticated AI. So all of us cobblers, factory workers, designers, technicians, service providers, and managers get our last few hundred years as necessary tools in the larger processes, perhaps not thinking quite enough of the long term term trajectory. The culmination of this process may stretch beyond the lifetimes of many of us, not unlike the potential outcomes of foolishness in our environmental stewardship. For now most of us may find ourselves living out our lives comfortably in the initially increasing, but then decreasing, number of service jobs available. With the rise of better, cheaper, more dependable tools, the global processes can only gainfully employ so many humans, and there’s every reason to expect that need to decrease.
This may come across as pessimistic or even fatalistic. I don’t think so. This is not a tale of apocalypse because, unlike much of even the best science fiction, I trust the creativity, ingenuity, and basic survival instincts of the human race to creatively adapt to the new circumstances. There will be change, and pain, as we come to grips with new challenges and devise new solutions. In situations where the change is very rapid the pain will be especially acute, and we may be caught unprepared, but change also brings opportunities. Remember the watch, the pressure of the minute hand? Remember the master of the schedule? We have spent so much time valuing and finding meaning in our lives through our careers and work. Getting kicked out of the factory may require us to reevaluate what it means to have a “meaningful life.” We may need to be a little less anxious about time, a little less divided between work and life, a little less possessive about what we’ve gotten “by ourselves for ourselves.” Simply put, we’re going to need something to do, something that gives us meaning and purpose. We need to find new ways to be creative, contribute to bringing order out of chaos, provide for our families, and be of service to others. It might not be too early to begin probing what these could be and begin preparing, both legislatively and socially, to help those who are unlucky enough to be employed in the first industries to become fully automated and heave the humans to the curb. Perhaps you can start by hugging a trucker.