AI will augment jobs in the future. But the argument could be made that job augmentation for some means job replacement for others. For example, if a worker’s job is made 10 times easier, the positions created to support that job might become unnecessary.
A June 2023 McKinsey report stated that generative AI would automate 60% to 70% of employee workloads.
AI is already replacing jobs, responsible for nearly 4,000 cuts made in May 2023, according to data from Challenger, Gray & Christmas Inc. OpenAI — the company that created ChatGPT — estimated 80% of the U.S. workforce would have at least 10% of their jobs affected by large language models (LLMs).
Examples of AI job replacement
One recent case of AI job replacement came when a writer at a tech startup was let go without explanation, but later found references to her as “Olivia/ChatGPT” on the work Slack channel. She also found communications from her managers about how ChatGPT was cheaper than using a writer. So, while there was no official explanation for the job loss, the signs pointed to AI.
The Writers Guild of America also went on strike, saying they wanted more regulation of AI in addition to higher wages and more residuals from streaming platforms.
Generative AI might also disproportionately affect the jobs of women, according to a recent study from the Frank Hawkins Kenan Institute of Private Enterprise. Approximately 79% of working women have positions that are susceptible to automation versus 58% of working men.
In past automation-fueled labor fears, machines would automate tedious, repetitive work. Generative AI is different in that it automates creative tasks such as writing, coding and even music making. For example, musician Paul McCartney used AI to partially generate his late bandmate John Lennon’s voice to create a posthumous Beatles song. In this case, mimicking a voice worked to the musician’s benefit, but that might not always be the case.
Other ways AI could affect jobs
Job replacement isn’t the only effect AI could have on work. The positive angle is human-machine cooperation. AI will help people improve their work experience by automating rote, repetitive tasks. The technology will maximize the “goods” of work while minimizing the “bads.” This may contribute to a surge in AI jobs and increased demand for AI skills.
AI needs a lot of human feedback. For example, LLMs train using a process called reinforcement learning from human feedback where people fine tune models by repeatedly ranking outputs from best to worst. A May 2023 paper also describes the phenomenon of model collapse, which states that LLMs malfunction without a connection to human-produced data sets.
There’s also a another angle — that workers will collaborate with AI, but it will stunt their productivity. For example, a generative AI chatbot might create an overabundance of low-quality content. Editors would then need to write additional content to flesh out the articles, pushing the search for unique sources of information lower on their list of priorities.
Jobs most affected by AI
Writing is just one example of a job being automated by the latest AI systems, but there are many job types that could be affected in various ways, including the following:
Administrative
Generative AI tools can help office administrators and assistants with tasks such as basic email correspondence, identifying data trends, finding mutually available meeting times across time zones and other summary/synthesis exercises.
Microsoft 365 Copilot — a collection of AI-powered tools integrated into Microsoft’s productivity suite — could radically increase office workers’ productivity. As of this writing, the tool is in beta testing and is expected to be released to the public later in 2023.
Coding
Programs such as ChatGPT can write fluent, syntactically correct code faster than most humans, so coders who are primarily valued for producing high volumes of low-quality code quickly might be concerned. Coders who produce a quality product might have nothing to fear, however, and use AI to improve their workflow instead.
Customer service
The customer service sector offers many opportunities for automation. AI-powered chatbots can provide speedy, personalized responses to customer questions, reducing the need for human workers. There are many examples of AI in customer service pre-ChatGPT, including the following:
- Robotic process automation.
- Customer self-service.
- Chatbots.
- Sentiment analysis.
It’s likely customer service departments will continue to integrate the newest AI technologies.
Currently, ChatGPT can’t automate an entire contact center, but there are many ways it could lighten the workload, such as translating or summarizing customer inquiries.
Legal
There is significant evidence indicating AI will affect legal jobs.
AI will eventually perform many of the tasks paralegals and legal assistants typically handle, according to one study by authors from Princeton University, New York University and the University of Pennsylvania. A March 2023 study from Goldman Sachs said AI could perform 44% of the tasks that U.S. and European legal assistants typically handle. GPT-4, OpenAI’s latest and greatest language model, passed the Uniform Bar Examination in the 90th percentile.
AI could help automate routine law tasks such as the following:
- Document review.
- Contract analysis.
- Legal research.
- Searches for relevant case law.
Teachers
Teachers could be affected by AI in several ways. The immediate concern is that they will have a harder time detecting plagiarism or students cheating on assignments. But AI could help teachers by doing the following:
- Acting as productivity tools.
- Drafting lesson plans.
- Generating quiz questions and mock tests.
Teacher sentiments range from being worried about the technology replacing them to insisting that the in-person classroom connection is essential to education.
Some people draw an analogy between ChatGPT and when students weren’t allowed to use calculators in the classroom. Now, most people have a calculator app on their smartphone. There might also be a time when it becomes accepted for students to use ChatGPT to aid with schoolwork.
Finance
AI is also making an impact on finance and banking. Generative AI could be used to monitor transactions and give detailed financial advice on how to save and spend efficiently. For example, Morgan Stanley uses AI-powered chatbots to organize its database.
Finance, banking and fintech have been invested in advanced AI for years. A 2020 survey from the World Economic Forum found that 77% of respondents in the financial services industry believed AI would possess very high importance to their industry in two years — research that lined up with ChatGPT’s release date.
Graphic designers
Adobe Photoshop’s new Generative Fill feature is one example of the way generative AI can augment the graphic design profession. The feature lets people with no photo editing experience make photorealistic edits using a text prompt. Other tools — such as Dall-E and Midjourney — also create realistic looking images and detailed artistic renderings from a text prompt.
Engineering
Generative design uses AI to expedite the computer-aided design process. Generative design helps with ideation, generating all computationally possible solutions to a problem within a given set of parameters — even when the design is completely novel and a radical change from anything that has come before.
Even though generative design affects the field of mechanical design, it is unlikely to replace human engineers.
Human resources
The hype around AI and the fear of job losses has created a difficult dynamic for HR departments to manage in the interim.
“[Generative AI] will infiltrate every aspect of HR,” said Patrick Thibodeau, a senior news writer at TechTarget, in a podcast. However, AI-powered recruiting tools could be used to do the following:
- Source qualified candidates.
- Review resumes.
- Automate recruiter tasks.
AI chatbots could also be used internally to help employees access their benefits and perform other self-service tasks.
What AI legislation is there?
In recent months, leaders in the AI industry have been actively seeking legislation, but there is no comprehensive federal approach to AI in the United States. Still, there have been state- and city-level rules passed. Several states — including California, Illinois, Texas and Colorado — have introduced or passed laws focused on protecting consumers from harms caused by AI.
New York City passed a bill, Local Law 144, regulating the use of AI in recruiting effective July 5, 2023. The law requires independent audits of AI hiring tools and publication of findings.
Google, Microsoft, OpenAI and Anthropic launched the Frontier Model Forum on July 26, 2023. The forum aims to advance AI safety research, identify best practices for AI development and collaborate with policymakers to develop applications that help solve societal challenges. Amazon, Meta, Google, Microsoft, OpenAI, Anthropic and Inflection also pledged to subject AI systems to third-party testing before releasing systems to the government, and to clearly label generated content.
The European Union has the AI Act, which will be the world’s first AI legislation and is expected to take shape late this year. The U.S. Congress is not likely to pass comprehensive regulations similar to the EU legislation in the immediate future. On June 21st, Senate Majority Leader Chuck Schumer formally unveiled an open-ended plan for AI regulation, explaining that it could take months to reach a consensus on a comprehensive proposal. Schumer emphasized that the regulations should focus on protecting workers, national security, copyright issues and protection from doomsday scenarios.
On Oct. 30, 2023, President Joe Biden signed an executive order on artificial intelligence. The order, though not a law, is a binding government action.The executive order aims to protect consumer privacy, create educational resources, create new AI government jobs, advance equity and civil rights in AI in the justice system and support workers in response to AI’s effects on the workforce.