Learning from the Luddites

Technology and the Future of Social Policy

The current discussion about robots and AI taking jobs from human workers is only the most recent example of technological progress causing workforce change. This is a continuous process that’s been happening to society since at least the invention of the wheel. Occasionally, however, we see specific instances where one type of technology accelerates change so rapidly that workers become uncertain as to whether they will be able to adapt. The issue here is larger than the technology itself; just as important is the economic and social process that will be necessary to keep the workforce happy, healthy, and stable.

You’re probably familiar with the Luddites: they were skilled, well-paid textile workers in Britain who saw their jobs being directly replaced by machines in the early 1810s that could produce fabric much more efficiently. It’s an apt historical analog to what robots are doing to some fields right now. The term “Luddite” has come to mean “a person opposed to new technology,” but that’s an unfair characterization of the Luddite movement. In fact, the Luddites were accepting of (or at least resigned to) new technology, and simply wanted manufacturers to use machines in accordance with standard labor practices. In an era of severe union repression in Britain, the Luddites used what has been referred to as “collective bargaining by riot” to seek assurances from employers that the machines would be operated by workers who were well trained and paid fairly, but the British government brutally repressed the movement.

The more recent past has seen fewer riots, but technological improvement continues, and so does its effect on workers. For example, new technologies like CNC machines have been making factories much more efficient, but different skills are needed to operate these machines than the previous generation of machines that were run mostly by hand. This has resulted in a situation of simultaneous high unemployment and high demand: some 600,000 manufacturing jobs are currently waiting for workers with the right skills and knowledge to fill them. It’s not as if workers are unaware of this skill gap, and some of them, at least, are trying to reeducate themselves. But it’s a time-consuming and often expensive process that’s not necessarily a realistic option for most people.

Both the Luddites and manufacturing workers are asking for reasonable things: namely, recognition of their value, and the opportunity to maintain that value in the face of rapidly emerging technologies. In the past, technological change has always worked out in the end for society in general, improving production, wages, and overall standard of living. But two questions need to be considered: how can we move forward more responsibly to help workers in the short term, and in the long term, is this robot-driven technological shift going to be different than the ones that have come before, as many have suggested? Will workers in occupations that are particularly vulnerable to automation be able to survive this transition?

Some people believe that robots and AI will reach a technological singularity in the next 20 to 40 years, where exponential improvement in the capabilities of both physical robots and software will virtually extinguish the market for the human labor necessary to fulfill the basic needs of our species. There’s also the suggestion that semi-skilled human labor facing a robot-driven economy is analogous to horse labor in the early 1900s, as pointed out by economist Wassily Leontief. At the beginning of the 20th century, 21 million horses were working on farms and in cities in the United States. The introduction of the engine led to a tipping point where buying and maintaining one mechanical horsepower became cheaper than buying and maintaining one actual horse, and horses lost their value nearly completely. By 1960 the horse population was down to just three million. When the cost of buying and maintaining one robotic “humanpower” becomes cheaper than hiring and providing pay and benefits for one actual human, the same thing could happen to human workers.

Maybe this is our future, but it probably isn’t. In addition to having history on our side, there are several points that need to be made about how robots will integrate into our economy. The speed and precision with which robots can complete tasks can seem daunting, but most roboticists are familiar with Moravec’s paradox, which is the inverse relationship between the complexity of the skill and the computational resources required to execute that skill. Robots excel at high level reasoning, but the basic sensorimotor skills that humans take for granted are still an enormous challenge for even the most sophisticated robots. This isn’t a characteristic that’s going to change anytime soon, giving humans plenty of time to adapt as robots (very) gradually become more capable. For the time being, “man is [still] the lowest-cost, 150-pound, nonlinear, all-purpose computer system which can be mass-produced by unskilled labor,” as NASA stated in a report from half a century ago.

In general, robots in industries like manufacturing can be seen more as tools than as complete replacements for workers. Due to their limited specialized skills, the robots usually can’t fulfill all aspects of a task on their own. Paired with a human worker, a robot can make the human more efficient while also improving working conditions, often making the human’s job more fulfilling by minimizing the actions that are ‘robotic’ in nature. And isn’t this what technological economic progress should be about? “As a species,” asks Rodney Brooks, “shouldn’t we get rid of jobs that are that dull and mind-deadening?”

Whether it’s an abrupt technological shift or a gradual one, we can agree that robots and automation are certainly going to alter our society, and even if it’s positive change in the long term, in the short term, it’s very likely that existing workers will see significant changes to their jobs. History suggests that we’ll be able to adapt, but what we’ll adapt into can’t possibly be predicted, which is why the way things are trending feels so concerning right now. This has always been the case, however, and the fact that we don’t know what’s going to happen in the next ten years, or even in the next five, should simply be recognized as normal.

This is not to say that there aren’t tangible, proactive steps that we can take to help ensure that this transition, however long it takes, happens as positively as possible. Essentially, the Luddites had the right idea, even if they went about it rather violently: the protection of workers through responsible and inclusive economic and social policies. To prevent humanpower from being treated like horsepower, all it takes is an active awareness that we humans are in control of our own economy, and we can collectively and globally decide to make and enforce decisions that are in the best interests of both the people who consume goods and services as well as the people who produce and provide them.

We should consider things like immigration reform to allow skilled workers to easily move to where work is available, or government grants for starting new businesses that take fair advantage of available human labor. Additional funding for basic research could result in entirely new industries, creating jobs where none existed before. And for the current generation of workers, substantial tax credits for companies who offer transition training, as well as special scholarships for those who need to become newly skilled could support workers and allow them to evolve with the economy. Lifelong education should be a significant part of the future of work.

It’s also important to recognize that increasing automation can contribute to a growing wealth gap, with the few people in control of the automation becoming extraordinarily wealthy. This is where social change becomes very important: recognizing that things like education, healthcare, and perhaps even a basic income should be subsidized by the profits that the workers at the heart of the economy are instrumental in generating. Once all workers are guaranteed the same benefits and protections that full time workers currently have, a creative ‘gig economy’ might become the norm for many people: supplementing their income how they want and when they want by doing the kind of jobs that humans find exciting and fulfilling and that robots find almost impossible.

This ideal represents a significantly different society than we have now, but it’s also very much a choice that we can make. We control our own society and our own economy, and we shouldn’t be afraid of accepting the increasing role of technology in our lives, as long as we do it fairly, and on our terms.