Will there be jobs available in a world of extreme robotization? We may not want to work, but work has been necessary for income, and thus consumption. If there are no jobs available, then most people will be very poor, dependent on handouts from the rich. That has motivated proposals for a universal basic income. But poverty is not the likeliest path. To get a better perspective, let’s imagine a science-fiction-like world in which our robots can make almost anything we ask them to make.
Mary wants a new bicycle. It requires a variety of components: frame, handlebars, wheels, and so forth. The components will be assembled by a robot, and then Mary will be riding through the hills. But the components need to be made. At some other facility, a frame is made from steel or carbon fiber, wheels are assembled from rims and spokes and tires, and so forth with all the other components. All assembly is robotic. The cost of producing the bicycle includes no labor, just energy and raw materials. Currently, labor constitutes the majority of the cost of most goods and services we use.
If the economy lacks competition, then the following analysis does not apply. But if one entity takes over all production, it doesn’t much matter whether it’s a robotic world: everyone will be subject to the decisions of the one entity. So let’s assume that even after the robotic revolution, companies still compete for sales. Entry and exit into various markets will be open, and any entrepreneurs who want to buy their own robots will be able to.
In a competitive market, prices usually fall to the cost of production, plus a return on investment that is commensurate with risk. The price of Mary’s bicycle need not compensate anyone for labor, but it must still compensate the owners of the other resources used. One might imagine that the companies that own the robots would rule the roost. But if a robot-owning company jacks up the price of goods that it makes, then another company will buy robots and undercut the price.
What about the cost of energy? In a sufficiently advanced world, robots can build solar panels or nuclear plants or hydrogen fuel cells. Robots can also drill for oil and natural gas, mine coal, or build dams. Energy will be extremely cheap, with the option for it will be clean or dirty.
What is left to impose a cost on Mary’s purchase are the natural resources: the steel, aluminum petroleum and raw materials that we use to make finished products. In today’s economy, someone owns the land or mineral rights, which are quite valuable. However, mineral rights will be exceedingly less valuable in a robotic world. Today a high grade iron deposit has greater value than a low grade deposit. But that difference is because less effort is required to get the finished iron from a high grade deposit: less labor, less machinery, less energy. In a robotic world, the lower grade deposits can be worked as easily as the higher grade deposits. It might take more robots, but robots made by other robots will be cheap. Most of the added value from high grade deposits will be lost.
The cost that would be left is scarcity value of natural resources. Scarcity value does not apply to robots, which can be produced, nor to the finished materials, but to the natural resources themselves. Planet Earth has only so much copper in its crust, for example, so the owners might demand a high price. If one owner sells cheaply, the other owners will smile knowing that their resources will be more valuable—because of higher scarcity in the future, as demand grows with population and available supply is used up.
This pattern has not worked in our world, though, despite fairly compelling logic. For all minerals, prices have fallen over time (after adjustment for general inflation). With falling prices, the resource owner prefers cash today over selling the resource at a lower price in the future. Some people will certainly speculate on higher prices, withholding supply from the market, but this is unlikely to be too common, given our history of falling natural resource prices. The scarcity theory does not work out very well in practice for two reasons.
The first challenge to the theory is substitution. Copper, for example, is not the only conductor of electricity, nor the only decorative metal, nor the only long-lasting sheath material. If copper prices rise too much, other materials will be used. They may not be quite as good for a specific use, but can be made to work. For example, most metals conduct electricity, including silver, aluminum, iron and gold. Other metals are suitable for jewelry. Copper has been used for plumbing and cooking, but other metals—and even plastics—are suitable for many of these uses. No resource price can climb too high, or users will shift to other products that provide the same functionality at lower cost.
The second challenge to resources prices is technological change. The current fracking boom in petroleum illustrates how productivity can be boosted. In effect, a given geological deposit now produces more oil than it would have a couple of decades ago. Virtually all natural resources have enjoyed technological improvement in their discovery, extraction and refining technology. We cannot say exactly what technological improvements will come in the future, but it’s certain that scientists and engineers will come up with innovations that make resource production cheaper.
Also note that the cost of a mineral today depends not only on the scarcity of the raw material but also the cost of excavating and processing the ore. Those costs will be very low in a robotic world, even though the basic material is still scarce.
Finally, another source of resources will be developed if and when scarcity pushes resources prices up: our landfills. We know that recycling is far from total today. At some point, engineers are likely to look into options for extracting valuable resources from previous generations’ discards.
So goods and services will be pretty cheap in a world of extreme robotization, but they won’t be free. How will people earn a living, even if the cost of living is incredibly low? The first people we think of are engineers who develop better robots, better processes, better machine learning techniques That’s an easy call.
Artists will likely do well, even though artificial intelligence is making great strides at creating beautiful images. I heard of an excellent potter whose work was sometimes confused with factory-made ceramics—so he intentionally made smudges on the bottom of his cups and bowls. Many people would like hand-made products, so a market will exist for artistic paintings, sculptures, jewelry, clothing, and household objects. Creators of music and literature will also be rewarded.
Some jobs will be very difficult to automate: moving a sick patient from one bed to another, for instance. And this leads to the most important conclusion. We will not reach extreme robotization all at once. The early robots solved fairly easy tasks that needed moderately paid workers. The first robots did not replace minimum wage earners, but middle-class, unionized workers. In the coming years, engineers will prioritize robots that are relatively easy to make, which replace relatively well-paid or hard-to-find workers.
At some point, the remaining tasks will be very hard, and wages for workers will be fairly low, and robotization will cease or at least slow down. That sounds bad—low wages—but remember that goods and services will be incredibly cheap when made in a roboticized world. Most likely, the cost of having employees will fall across all skill levels. (If one job skill does not fall in wage rates, the engineers will concentrate their efforts on that occupation.)
Look for a world with cheap goods. There will be jobs available—those that are hard to robiticize. With the cost of living low, many people are likely to choose more leisure and less work. Our challenge will not be supporting ourselves and our families, but rather finding meaning in a world in which less work is needed.
