Photo by Steve Johnson / Unsplash
AI and the Labor Market: Complement or Competitor?
In only a few years, artificial intelligence (AI) programs like ChatGPT have gone from being impressive curiosities to widespread disruptors of industries. A major concern is that many workers will be replaced by AI programs, which can compile and process tremendous amounts of data and provide a human-esque answer. Additionally, AI can often successfully interpret input prompts by laymen (non-experts) and create sophisticated outputs, reducing the need for specially trained users.
Abilities of AI
AI has largely created “fuzzy logic” for computers, allowing laymen to access much greater ranges of computing power. For example, casual users can ask popular AI programs, such as ChatGPT, to create schedules and charts compiling data from multiple sources. Several years ago, this would have required multiple programs and lots of time. Today, ChatGPT can do it in less than a minute. Programs like ChatGPT can search the Internet for many sources of information faster than human users and summarize that information into definitive answers, like “yes” or “no”. It can link specific pieces of information in its answer to those sources, allowing the user to ascertain their reliability.
Controversially, AI can format responses to questions according to user desires, including formal papers meant for academic purposes. This has led to fears of a wave of cheating in schools as students ask ChatGPT to write specific-length and specific-format papers about various topics. Some research, however, finds that AI did not significantly increase the incidence of cheating after it was unveiled in 2022. Still, critics argue that AI can now create the entire finished product, such as a formal paper, while previous cheating required the student to at least generate some of his or her own work.
Generative AI
Generative AI allows the creation of images using only text as input, with machine learning allowing the program to find and analyze thousands of similar images to create a new, unique version. Instead of hiring artists, advertisers could theoretically use generative AI to create their own ads, both in still images and in video clips. Instead of relying on receptionists and assistants to craft communications, generative AI programs could rapidly create high-quality responses, including answers and sourced information, that directly answer customers’ questions.
Economic Implications: AI as Complement or Substitute (Competitor)
The world is currently abuzz with AI’s upcoming effects on labor markets, especially in developed economies currently saturated with AI-using technology like desktop computers, laptops, tablet computers, and smartphones. Proponents of continued expansion of AI access and usage argue that AI will mostly serve as a complement to labor, enhancing demand for labor and increasing workers’ wages. Critics of AI expansion, however, argue that AI will mostly serve as a substitute to labor, decreasing demand for labor and reducing workers’ wages.
AI as Substitute: Entry-Level (Non-Executive) Jobs, Simple Physical Skill Jobs
Most observers expect jobs with less “deep thinking”, human connection, or complex physical skills to be automated first by AI programs. Already, many jobs that require workers to find and report simple pieces of information, such as personal assistants, data entry clerks, receptionists, and customer service representatives, are being automated by AI chatbots. These positions are unlikely to disappear entirely, but only those who are most senior (in terms of tenure) or skilled are likely to remain as overseers of the AI program(s). For example, an office with six clerks may end up with only two clerks to oversee various AI programs running the office’s calendar, data entry, billing, communications, and social media ad posting functions.
Jobs requiring simple physical skills, such as manual labor in a fixed location like a factory, are at increased risk of being automated by AI. New AI programs will likely increase the ease of using existing machines to replace workers for simple tasks. They could also automate and “smooth out” the intricacies of using specialized equipment, allowing lower-paid workers to use the equipment with roughly the same results as experienced workers. Essentially, if a machine currently exists that can automate a job, this trend will likely accelerate due to AI.
Warning:
Even non-executive jobs that require advanced knowledge, such as computer programming, language translation, and expertise with specific technical vocabulary (such as medicine or law) are highly vulnerable to AI. Currently, AI can generate computer programming code at near-expert levels, allowing non-programmers to craft their own software using AI. AI can also swiftly glean information from reams of dense medical and/or legal texts, also allowing laypersons to come up with correct answers that were, until recently, left to medical clerks and paralegals. In these more skilled fields, companies may replace higher-paid programmers, translators, paralegals, and medical clerks and transcriptionists with lower-paid laypersons who operate AI programs.
AI as Complement: Executive- and Expert-Level Jobs, Complex Physical Skill Jobs
Workers who make executive decisions are less likely to be automated, especially when their decisions could impact health and safety. For liability reasons, there are many jobs that governments and voters will likely not allow to be fully automated, at least in the short term: law enforcement, medical, in-person education, utilities, etc. While AI programs will almost certainly come to assist these endeavors, the public will want human workers on the job to protect against any errors. After all, would the public trust AI entirely with arrest powers, medical treatment, overseeing the safety of children in school, or handling dangerous materials?
Executives who provide orders and oversight to front-line workers will also not likely be replaced by AI, as there would be strong worker resistance to taking orders solely from an AI program. Those who are responsible for managing others often have to deal with complex interpersonal situations, behavioral nuances, and moral “gray” areas that AI has yet to master. Managers and executives could use AI to find examples of how worker behavior X was dealt with in the past by similar firms under similar circumstances, but the public would likely not accept decisions like hiring, promotion, demotion, and firing to be done solely by AI.
Jobs requiring complex physical skills, such as construction or machine operation requiring moving from place to place, are less likely to be automated by AI. Because AI is software rather than hardware, jobs that cannot currently be automated with machines are likely to be secure in the short term. A construction worker who has to move swiftly around a construction site with multiple tools will be difficult to replace, as opposed to a stationary worker on an assembly line.
Outlook for Workers: How to Protect from AI Outsourcing
AI may quickly remove the wage-boosting effects of developed knowledge, meaning workers should focus more on developing interpersonal skills, complex physical skills, or true “expert” status rather than deeper knowledge like learning a new language or memorizing data. Historically, fears of mass unemployment from automation have been overblown, with disruptions in specific industries only being short-term or isolated. While specific jobs and industries may be heavily affected by AI, workers who are forward-thinking and open to exploring the rise of new careers will likely find new employment quickly.
The new efficiencies and increased output generated by AI will require workers in new areas to process that output, so all workers should follow the latest news and trends on where this output is being generated. Developing a working knowledge of AI programs and how to successfully use AI prompts will be very beneficial, as firms will look to retain those workers who can oversee AI tools. Not all workers vulnerable to automation will be replaced, but only those who are the most tech-savvy and have the best people management skills will likely be retained in those current jobs.
Summary: AI is Both Complement and Competitor, so Work on Your Interpersonal Skills!
AI is good at replacing the technical know-how and retained knowledge to do a task on a computer, so everyone should focus on bolstering their interpersonal skills. Executives and experts will likely keep their jobs, and should focus on using the time freed up from online research and manual communication-crafting to develop strong relationships with colleagues and clients. Entry-level workers are vulnerable to automation, and so should also focus their freed-up time to improve their interpersonal skills so they will be seen as the new “managers” of AI programs that handle communications with other workers, clients, and companies.
