ChatGPT and Work – Will Generative AI Replace Your Job?

There’s been an avalanche of news about ChatGPT and work – how it could result in massive job losses among office workers and the educated class. Ironically, after watching traditional industries collapse in terms of employment (but not production) in the previous century, the same fears now rise among white-collar workers.

Need a reminder? In 1960, the steel industry in America employed 700,000; today, it’s dropped to just 83,000. Productivity went from 10 person-hours to produce a ton of steel to just under two person-hours today. 1.4 million were employed by the railroads in 1950; today, it’s just 177,000 people who are 24 times more productive in terms of the amount of tonnage moved. Areas that were once economic powerhouses were decimated, with cities such as Buffalo, Cleveland, Detroit, and Pittsburgh losing more than 40 percent of their populations. Many only began to recover in the past decade.

ChatGPT and Work

ChatGPT and Work - from ChatGPT's perspective.

Put aside the scare headlines, ChatGPT has the potential to take over the work of millions of people. Asking what tasks it could do (on March 9, 2023), it created a helpful list, though with its usual reassurance that “. . .  my main goal is to assist users in any way possible.” AI may be the tsunami in our lives this decade, but it wants everyone to stay calm. So here is its list of the parallels between ChatGPT and work:

  1. Answer questions: I can answer questions on a wide variety of topics, from science and technology to history and current events.
  2. Generate text: I can generate text on any subject, from short paragraphs to longer essays.
  3. Provide recommendations: I can recommend books, movies, TV shows, and other forms of media based on your interests.
  4. Translate languages: I can translate text between different languages.
  5. Assist in writing: I can help with writing tasks such as generating ideas, outlining, proofreading, and editing.
  6. Engage in conversation: I can engage in conversation with users and respond to their questions or comments.
  7. Provide information: I can provide information on a wide variety of topics, such as weather, news, and sports.

If you’re in a job that involves sales, writing, generating content, engaging in conversation, or office-based work, you definitely have cause to worry. And a few other careers it didn’t bother to list, such as software programmers.

A Massive Increase in Wealth Disparity?

By the end of 2021, the top 1% of households in the United States held 32.3% of the country’s wealth, while those in the bottom 50% held only 2.6%. An article in The Guardian notes that income inequity may be the most significant impact of the AI revolution.

ChatGPT is just the latest technology to fuel worries that it will wipe out the jobs of millions of workers, whether advertising copywriters, Wall Street traders, salespeople, writers of basic computer code or journalists.

But while many workforce experts say the fears that ChatGPT and other artificial intelligence (AI) technologies will cause unemployment to skyrocket are overblown, they point to another fear about AI: that it will widen the US’s already huge income and wealth inequality by creating a new wave of billionaire tech barons at the same time that it pushes many workers out of better paid jobs.

Of course, new jobs will be created by the rapidly expanding use of ChatGPT and other AI platforms. That’s true of any tech development. But the challenge – as it was with workers in traditional industries in the last century – is whether or not people can retrain for AI careers and how many new jobs will be created.

There was a sad joke that made the rounds in the 1960s and beyond – can a 40-year-old steelworker retrain to be a nurse? It could easily be worse this time around – what exactly will a 40-year-old office worker retrain for?

ChatGPT and Surveys of Employers and Employees

ChatGPT and job loss studies.

If you’re looking for evidence of who gets their job axed by ChatGPT and who is safe, it’s too early in the conversational AI revolution for any comprehensive studies. The only study we could find was a recent one done by SortList, Europe’s largest B2B matchmaking platform. They surveyed 500 employees and employers users in 6 different countries to see if they saw ChatGPT as a threat or an opportunity. We’re sure there will be better analyses in the future, but we have to go with what’s available now.

Some quick highlights from the SortList study on ChatGPT and work:

  • 43% of employers want to hire ChatGPT as a copywriter.
  • 26% of employers plan to cut jobs in software and tech.
  • 51% of employers are looking at ways to automate marketing through ChatGPT, however, only 16% of employees in those departments fear layoffs.
  • As for those working in education in the corporate sector, employees are twice as likely to fear job cuts than what employers predict will happen.
  • Not surprisingly, Millennials in both tech and finance are 2.4x more likely to worry about ChatGPT taking their job.

Here’s one chart from the study that lays out the differences between employer predictions and employee fears:

employer predictions of job cuts and employee fears of job losses due to ChatGPT

Too Early To Predict The Job Losses

Some jobs will change dramatically as we enter this new world of augmented intellectual activity. And some careers will be decimated, especially in areas where AI can push out content or chatbots can replace live interactions with humans. What we tend to forget is that the impacts here will be felt around the world. Most companies have already outsourced their call centers to countries with lower wages. While those jobs are not the best, they’ve often been stepping stones for people in developing countries to move up the economic ladder.

Other careers will see deep integration of AI, which will result in job losses and change how people work. Software developers won’t all lose their jobs, but they will become integrators and facilitators as ChatGPT helps them do routine coding tasks. With programmers, you might see a parallel with what happened in the construction industry with the introduction of heavy machinery. Construction workers haven’t disappeared, but it now takes many less to construct much larger buildings as machines have augmented their work.

As For Education . . .

There are wild predictions that ChatGPT will replace K-12 teachers and university faculty. Honestly, conversational AI has a way to go before we reach that point – though everything depends on the pace of progress in AI. We should expect massive job losses in support staff in K12 and higher education in areas such as academic counselors, tutors, and even librarians. AI that can accurately answer questions 24/7 will quickly put these positions in jeopardy. And it will begin to replace teachers in remedial courses in the next few years where content is standardized and students are learning basic skills.

If you’re in doubt, look at the Kahn Academy with its more than 8,000 videos and 7.7 million subscribers on YouTube. It isn’t AI; it isn’t even an interactive platform. But students flock to it for tutoring, test prep (SAT, LSAT, etc.), and academic support. Why would students wait until someone shows up on campus (or on Zoom) during the day when they can interact with human-like chatbots that accurately answers their questions and guide them through their academic work?

Generative AI and ChatGPT will profoundly impact the workforce in the future. And no one – not even OpenAI – has a solution on how to manage the disruption. As The Guardian article reminds us, the tech revolution has already created a billionaire class and pushed a significant segment of the population down the socioeconomic ladder. Expanding that gap will have a far greater impact on our social and political stability than the loss of careers and the augmentation of others through AI.

Let us know your thoughts as we plunge into a new era where AI will be our work colleague. The main question will be: what skills can you offer – or acquire – that a machine cannot replicate?