I, Unemployed: Why AI Has High Potential To Be Devastating To Workers Of Every Kind
When AI is utilized as a replacement for workers, not a tool in their tool belt, how will the average person be affected?
Author: Patrick Hart
Photo by: Patrick Hart
Across the breadth of human history, there have been many technological advances that allowed us as a species to flourish beyond life as nomadic tribes. The wheel afforded humanity the ability to transport large amounts of goods and ourselves in fractions of the time and effort. Irrigation brought us agriculture, so we could settle and, by proxy, create complex societies, security, and art. The printing press led to the democratization of information, allowing what we know as a formal education to be more accessible to those less well off. And the industrial revolution was the start of our ability to mass produce goods, eventually bringing prices down for mass consumption.
On paper and long after the fact, these advances are net wins for humanity once we eventually learn how to integrate our workforce in and around them. The assembly line, for instance, allowed us to rapidly expand as a species, but since the industrial revolution, these advances have often come at great cost to a large minority. For example, the automation of the assembly line was a leading cause in the mass unemployment downfall of one of the United States’ most historically prosperous cities; Detroit, Michigan. Such advances over the course of time are good for our continued growth, however, when implemented prior to governmental contingencies being created to protect its workers and citizens from the inevitable dropoff in employment, these advances spell disaster for many people.
These examples are important because it is paramount to look at the past when assessing the future. AI and robotics, like any automation technology, comes at a heavy price for those that make their living in jobs that repetitive tasks are the largest part of; and with these two advancements specifically, the number and type of jobs at risk are far beyond anything we have encountered before. This assessment can be done across most industries, but let’s look at the book publishing business, an industry I myself am working toward entering with my furthered education. Once AI reaches a threshold of intelligence deemed competent enough, automation will not only include low-level clerical work or warehouse operations, but jobs up until recently only humans functioning as higher level publishing staff could take on, but we’ll get to the specifics on that later.
With statisticians already theorizing an astronomical ~400-800 million jobs lost to the implementation of AI by 2030, it’s showing itself to be an ever present threat to most industries and the human beings that rely on that income to live. Due to the exponential improvement in the function of AI and robotics, and the consistent iteration in new models of each that take on different kinds of tasks, that already large number could change to be far greater than expected. The hope, of course, being that the statistics will be wrong by a large margin in the working human’s favor. If left unchecked, using book publishing as the example, given that AI models already exist that tout things like the ability to create, edit, and publish novels through their platform, we are looking at a potential repeat of what happened in Detroit en masse, where the already diminishing middle-class creates an even larger divide between the rich and the poor. Think like a CEO for two seconds and realize that creating, editing, and publishing are the bulk of the job of a publisher, with the added benefit of generative AI to replace the writer. So, with one program, you, as CEO, can replace the majority of your workers in one fell swoop.
With the average working age population at roughly 3.5 billion people worldwide per year, a layoff statistic of up to ~800 million has the potentiality to put the overall unemployment rate of working aged people in 2030 at ~22.85%; that is ~16.34% higher than the recorded high in the worldwide unemployment rate of 6.59% during the pandemic in 2020 where ~230 million people were out of work.
From here on out, I will take you through the frontline jobs in the publishing industry and explain how not even seasoned veterans of the trade will retain job security in the coming years if the corporate desire for rapid profit growth exceeds a humanitarian desire to take time to implement these technologies in an ethical manner to stave off economic disaster. Before that, let me quickly first define what AI models are at their basest root. AI models are programs that utilize artificial intelligence trained on immense amounts of data to analyze data, establish and recognize patterns, and solve queries on their own with minimal to no human interaction.
Warehouse Workers and Drivers: The physical pivot point of the publishing business, books need to be moved from one point to another, stored and moved for sale in both a business-to-business (B2B) and business-to-consumer (B2C) capacity.
Why is it at risk? With the advancements in AI and robotics, companies such as Amazon are beginning to shift to a robotic warehouse model. Warehouse work being repetitive in nature, moving boxes from one place to another, and boxing up goods for shipment, means this replacement of human workers and the implementation of these technologies is relatively simple. As for drivers, automated driving continues to improve year after year, utilizing increasingly optimized AI to read maps, roads, driving conditions, and assess immediate influences in the form of traffic with the implementation of cameras placed in, on, and around the vehicle. This is still some years off, though Tesla’s implementation of EV Semi’s is happening eventually, and with Tesla brand vehicles comes the software and technology to be self-driving, however the self-driving of vehicles is as of yet not legal in every state or country. Once they are, the entire trucking industry is at risk, regardless of what goods they are carrying.
Customer Service Representatives: The foundation of commerce, the customer-facing workers who take care of queries on any number of topics; general customer questions, book orders, release dates and availability, and questions regarding customer accounts.
Why is it at risk? This is already being phased out across many industries in online and phone formats. The customer service representative job is a series of tasks focused almost exclusively on data analysis, problem solving, and resolution. The bottom line of a corporation's bottom line is that it is completely serviceable by what current AI models are capable of doing to an acceptable (to them) degree. I’m sure I’m not the only person shouting into the phone when getting the pharmacy automated system that, “I’d like to speak to a human, please”, for fifteen minutes before being sent to a voicemail…
Copy Editor: The quintessential book editor position often solely tasked with editing an acquired book for grammar, spelling, and punctuation.
Why is it at risk? A copy editor’s job, in essence, is to scan a document and correct any problems in the text in accordance with a standardized ruleset laid out for writing in whatever country they work in and are distributing to. This task, with AI, can be automated easily, because it is foundationally data driven problem solving. If a publisher utilized an AI copy editor, each publisher-side edit would take seconds, leaving the bulk of the time spent in the revision phase tasked to the author…if there is one.
Production Assistant and Manager: These jobs entail everything from copy-editing, book and page layout/formatting, printer selection, and materials sourcing.
Why is it at risk? If a publishing company has a standardized format for its production process, an AI model could be sourced or created to automate the majority of these tasks, leaving potentially some room for more B2B interaction-based actions such as materials sourcing and printer selection to be done by a human. However, with the exponential rate of growth in AI programming, it is very possible these tasks could be automated as well, given that they are largely order and expense driven, especially if these B2B relationships are primarily online.
Copywriter: A fundamental part of the marketing team at any publisher, this job exists to craft digestible information about forthcoming books for both B2B and B2C marketing. This can include synopses for the back cover, tipsheet, onesheet, and email marketing campaigns, as well as any copy necessary for social media and event planning.
Why is it at risk? A copywriter’s primary task is to summarize large amounts of text into palatable bite-sized chunks. Currently, when one opens a document in Adobe Acrobat, with an Adobe AI subscription, they can ask the program to summarize the entire content of the document for them, it then provides them with a summary of the contents, and bullet-pointed lists of the most important takeaway information necessary for the reader to understand. With some tweaking, an AI model could be tasked with taking this idea, and inputting the company’s particular copy style into it.
Why should we be weary of all of this? I’ve only laid out the risk for a single industry, but businesses far and wide have jobs that fit the criteria able to be worked by AI models. AI can and will be a great tool for all industries and their workers eventually, freeing up human workers to focus on tasks far better suited for their time rather than on the repetitive tasks that eat up that time. Why we should be weary of this is fundamentally (due in large part to the nature of the way our socio-economic structures are set up) the way that corporations spanning all industries are looking at AI, as replacements for human workers to bolster profits, and we are not yet prepared for mass unemployment on this scale. Without a governmental initiative to supply its populace with both a Universal Baseline Income (UBI) and Universal Healthcare (UH), we will quickly see structural issues within the world economy as a whole. If the worst case scenario of 800 million unemployed people comes to pass, and they are struggling to make ends meet, that quick swell in profit margins will take a swift dive when a large percentage of the population is unable to purchase the goods and services businesses are providing. As the book publishing industry is concerned, since this was my example, books are a luxury, not a necessity when survival becomes the goal.
The argument should not be that AI will afford us an ability to make more money with layoffs because an AI doesn’t come with the annual cost of salary, healthcare, breaks, lunches, leave, and vacation. The argument, ethically, has to be that AI will eventually afford us the ability to enrich our workforce’s lives with the automation of repetitive tasks, research, and management. This argument can only come with an end goal of replacing these workers eventually, as long as the proper measures are taken to ensure the safety, security, and re-education of these workers into other avenues of employment.
I’ll leave you with this, when we first began nuclear research, it was to understand the nature of the atom and its potential use as an energy source; very quickly that research was used not to better our lives, but to end lives through the Manhattan Project. The utilization of AI and robotics is not such an immediate threat to our livelihoods, and as sardonically hilarious and sad as it is, our downfall by AI will not be so grandiose as science fiction so often makes it out to be. It will be slow, grueling, utterly mundane, and economical. No, there will be no sentient Skynet AI to blame, it will be because of greed. Those very real threats have the potential to be in the form of a massive uptick in the epidemic of homelessness, and the inability to take care of and provide for oneself and one’s family. So these industries, like the book publishing industry, can set themselves as examples of business to follow, where human workers come first ahead of the potential profits their AI replacements might dangle in front of them.
I encourage you to build on the research I have done here and find those ways in which you can help mitigate this potential disaster, adding more informed voices to the chorus never hurt.
References:
Novak, C., & Sander, L. (2021, December 1). Where Does Detroit Go From Here? A (2021) Review of The Origins of the Urban Crisis (1996). Midstory. https://www.midstory.org/where-does-detroit-go-from-here-a-2021-review-of-the-origins-of-the-urban-crisis-1996/
Manyika, J., Lund, S., Chui, M., Bughin, J., Woetzel, J., Batra, P., Ko, R., & Sanghvi, S. (2024, July 16). Jobs lost, jobs gained: What the future of work will mean for jobs, skills, and wages. McKinsey & Company. https://www.mckinsey.com/featured-insights/future-of-work/jobs-lost-jobs-gained-what-the-future-of-work-will-mean-for-jobs-skills-and-wages
Dyvik, E. (2024, July 4). Global employment figures by gender 2022. Statista. https://www.statista.com/statistics/1258668/global-employment-figures-by-gender/
O’Neill, A. (2023, April 4). Global unemployment rate from 2004 to 2023. Statista. https://www.statista.com/statistics/279777/global-unemployment-rate/
Holzer, H. (2022, January 19). Understanding the Impact of Automation on workers, jobs, and Wages. Brookings. https://www.brookings.edu/articles/understanding-the-impact-of-automation-on-workers-jobs-and-wages/
6. Amazon. (2024, February 20). A Day in the Life Working in Robotics at Amazon. Youtu.be. https://youtu.be/kxNkfW0OXcM?si=GyOQuQ97W7vveXSB
Comments
Post a Comment