Moving Past Automation Replacement Anxiety
Why fears of a labor apocalypse may be overblown and how automation could create more jobs.
As an ardent believer in automation, I utilize AI in my job. When I have the financial means, I invest in shares of automation companies. I fervently hold the belief that automation and AI will lead to a new industrial revolution. Now that I have acknowledged my bias, I aim to present an unbiased, yet optimistic counter-argument to address concerns about automation anxiety and the fear of a labor apocalypse.
Automation Before the 21st Century
Although automation has captured headlines in recent years, it is worth noting that we have been living with automation long before the advent of burger flipping robots, Open AI and Chat GPT.
Textiles and the Luddites
The Luddites are in vogue once again as automation continues to be top of mind for many workers. During the industrial revolution, the Luddites famously destroyed machinery in an attempt to stymie industrialization. However, this uprising was regretfully met with loss of life and suppression by the British government. In the end, the textile industry grew fifty-fold, creating many more jobs than the Luddites had initially sought to preserve.
The Electric Dishwasher
Some automations are so common that we often forget that they are, in fact automations. Commercialized in the 1950s, the dishwasher is now a fixture in millions of homes and restaurants. However, it is important to note that the job of a dishwasher has not disappeared entirely. It has enabled a single employee to handle dish washing while also attending to other responsibilities. In my personal experience, this meant taking on additional tasks such as mopping, bussing tables, and eventually moving to the front of house to make malts, as long as the dishes were clean.
Lego’s Warehouse Robots
Since 1987, Lego has been employing AGVs (Automatic Guided Vehicles) to transport boxes of finished products within their factories, ensuring smooth production flow. While this particular job has been automated, Lego still relies on humans to ensure that every brick meets the desired standards of perfection.
One Step Back, Two Steps Forward
The fear with automation is that it will replace some jobs, and there’s no getting around that fact, on some level, it will. Let’s consider Miso Robotics’ Flippy Robot as an example. This robot is priced at $36,000.
Considering that many states are either considering or have already raised the minimum wage to $15 per hour, employing workers to operate a grill at a fast food restaurant with operating hours from 6AM to 12AM could amount to a daily cost of $270. Extrapolating this over a year, the annual cost would reach $98,500.
In theory, by automating one station, the restaurant could save $62,500 annually. If the owner were to automate two stations, the savings would increase to $125,000 annually. While this number is not perfect, for the sake of argument, let’s assume this thread is correct and that eight employees work each shift, and let’s assume this restaurant has three shifts. This would mean six out of 24 jobs are affected, amounting to 25%, which admittedly is a large percentage.
Now, the restaurant owner has some options:
- Keep all the extra profits for themselves.
- Provide the remaining employees with a pay increase of one dollar per hour ($16/hr) and still have money left over. Since the remaining employees become responsible for overseeing the Flippy robots and thus contribute to increased productivity, they thereby have an economic right to receive higher pay.
- Open a second restaurant, creating 18 more jobs. The savings from both restaurants could help repay a $1 million loan in four years.
To me, the choice seems clear: open the second restaurant. By doing so, the owner would have created 36 jobs at the expense of 12.
It is worth noting that, prior to the notion that these jobs — such as burger flipping and food frying — could be automated, there was little desire to pursue them. Firstly, they involve inherent risks, such as the constant threat of burns and slippery floors. Secondly, growing up, I recall people using the phrase “flipping burgers” as a pejorative statement. Although I personally do not share these beliefs, having worked as a dishwasher and a pizza cook, it would be willfully ignorant to disregard the significant contempt American society has historically held for these occupations.
Reshoring Jobs
It’s no secret that the US has found itself at a crossroads with regards to globalization, Apart from the ethical concerns associated with exploiting populations to lower the prices of goods, the U.S. is contemplating if it’s even worth a return to the status quo.
In 2018, America initiated a trade war that has since evolved into something of a break up between the US and China from an economic standpoint. As companies search for alternatives, the possibility of reshoring these jobs has increased, albeit with a substantial emphasis on automation.
Beginning in 1978, the US started the process of moving jobs to China. Prior to this, the U.S. began moving jobs to Mexico. As late as 2011, hourly wages for international manufacturing looked something like this:
- United States: $35.53 USD/Hour
- Mexico: $6.48 USD/Hour
- China: $1.36 USD/Hour (2008)
- India: $1.17 USD/Hour (2007)
To Americans, the wages for Mexican and Chinese workers must seem criminally low. At this point, trying to undercut India and China would round to zero, and that’s precisely what I’m suggesting here. Amortizing the cost of automating manufacturing would eventually zero out for businesses. The cost of the entire workforce overseas could be redistributed to a handful of American workers being paid American wages.
Let’s consider this: Currently, these jobs are not American jobs. Bringing back any manufacturing jobs from overseas would be a victory over the current circumstances, even if many of the once-overseas positions are automated. American workers would now oversee the plants, operate and maintain the machines, which is currently not the case.
In the event that the U.S. successfully reshores manufacturing jobs, what’s to stop countries like China and India from automating entirely and repeating the cycle again? This could involve some persuasion from Washington, in the form of additional tariffs. There’s also the matter of the additional shipping costs to move products across the Pacific, whereas domestically those costs are mitigated. In truth, it’s a heavily nuanced problem to solve, which can’t be fully considered within the confines of this piece.
Just to affirm my position, I am in favor of globalization and free trade. I believe that free trade is a cornerstone of a healthy economy. However, I see an opportunity for the US to utilize automation as a tool to regain some leverage and create jobs at the same time.
The Environment
Every day, 172 vessels dock at U.S. Seaports adding to the 1 billion metric tons of carbon dioxide emitted by the shipping industry. The disruptions in the supply chain caused by COVID-19 and various other misadventures in the shipping sector have made our reliance on global supply chains a clear vulnerability. Reshoring a portion of these goods presents an opportunity for the United States to generate jobs while also reducing carbon emissions. While there may be some job losses in the shipping industry, this proposal appears to be mostly a win-win.
Ethically Driving Down Costs
The dirty secret behind American consumerism is that the affordable prices we enjoy are built upon the exploitation of vulnerable populations. Unfortunately, Americans are deeply dependent on this model, and recent inflationary pressures have revealed the fragility of this system. Localized automation, however, offers a solution by providing a fixed cost for manufacturing goods that gradually approaches zero or the cost of materials. Companies that adopt automation not only save on overseas labor and shipping expenses but also have the opportunity to amortize the initial investment over time. Consequently, this leads to more affordable goods without the ethical dilemmas that currently exist.
Benefits to the Economy
Satisfying the US economy’s need for exponential growth amid declining birth rates poses a challenging dilemma for policymakers. How can an economy continue to expand when there are fewer people available to fill existing jobs? Moreover, this doesn’t even address the need to create new jobs to sustain economic growth. The consequences of a shrinking GDP can result in a recession or, worse, a depression. By embracing automation, the US can supplement the workforce and mitigate these inevitable challenges, whether as a temporary measure until birth rates recover or as a long-term solution.
Doing the Invisible Jobs
It’s difficult to claim that automation is taking jobs when the jobs it’s doing today were never paid to begin with. Recently, in my role as a product designer, we needed imagery to enhance the UI of the product. There was no budget for these images, not a penny. Essentially, this was not a paying job to begin with. We had a few stock photo credits remaining on our account, but stock photos, by their nature, are commonplace, and we wanted something that felt like a unique, high-quality 3D asset. In short, this was an impossible task.
Hiring a 3D artist to develop these assets for us would cost hundreds of dollars and take weeks, and time and money were two things we didn’t have. Instead, I turned to Midjourney to generate these assets, and within minutes and for pennies, I had high-quality 3D models that I could use in the product. The choice wasn’t whether to hire a human or use AI; the choice was to use AI or not have the images at all.
The Dangers and Realities of Automation
Automation and AI are not without risks or consequences. Let’s explore the decisions that need to be considered and addressed before America adopts automation more broadly.
The Need for Regulation
I believe the economic potential of automation is undeniably enticing, almost intoxicating. However, if I know capitalism, and I do, there may be a temptation to automate all jobs. This is why we would require an automation tax that aligns with the level of automation adopted by businesses. The basic premise of an automation tax would be that automating a small percentage of jobs would incur no tax, but as certain thresholds for are surpassed, the economic advantages rapidly diminish, discouraging the foolhardy pursuit of automating significant portions of a business.
It is important to clarify that the stated model would solely apply to existing businesses. For new businesses that create new jobs (such as reshoring manufacturing), the ratio would be inverted since those jobs are not replacements but entirely new positions. This approach fosters economic expansion without gutting the existing workforce.
However, it must be acknowledged that this approach is not flawless. There is a clear loophole where businesses could close down and subsequently reopen to enable more automation. Additionally, existing businesses could argue that they cannot compete with emerging competitors due to the differing tax ratios. Perhaps the ratios could be adjusted over time, but I would believe it important to defend current jobs while also encouraging manufacturing to move state-side.
If the governing powers don’t assume the responsibility of regulating automation, it will be left to the courts, and by then, it may be too late. We are already witnessing this process unfold, but for this mechanism to even begin, there needs to be some harm done to trigger a lawsuit. While advocates of small government might consider this approach sufficient for regulating AI, relying on court cases is time-consuming and imperfect when it comes to addressing broader issues that are better suited for the Administration, House, and Senate.
The Need for Guidelines when Automating Tasks
Not all jobs should be automated. Take, for example, the disastrous crashes of the Boeing Max 737 that killed 346 people. In both crashes, the autopilot system malfunctioned, causing a tragic loss of life. Stepping back, the issue may run deeper than just bugs in an autopilot system. Although autopilot is not a new concept in aviation, the degree to which modern planes are adopting automation is alarming.
Simultaneously, self-driving cars have been under development and are now hitting the streets of San Francisco with mixed results. Although these cases seem closely related, they differ in one key respect: choice.
Airplane passengers have no say in the crew that flies their plane. They are beholden to the airline’s decisions to automate or not automate these jobs, and the limited options force flyers to either accept an automated crew or not fly at all.
Conversely, self-driving cars and taxis face a multitude of competition. If a prospective rider is unwilling to accept the risks of a self-driving car, they can simply hail another cab or, in the case of piloting a self-driving car, deactivate the feature and take control. The primary difference comes down to the choice of the end-user and whether that person is willing to accept the risks that come with automation.
Summary
Along with its enormous benefits, automation carries significant risks. The advantages of automating our economy include job creation, reducing carbon emissions, offsetting population declines, strengthening our supply chains, and ethically reducing costs. However, without regulation and guidelines for the practical and ethical implementation of automation, the potential damage could be dire and irreversible.
All image generated with Midjourney, scaled & expanded with Runway ML. Concept & retouching by David Kelly.
