AI Will Change Your Job. Can You Change Along With It?
Unions such as the Writers Guild deserve credit for leading the discussion about the best ways to address the risks posed by artificial intelligence.
Artificial intelligence is going to eliminate jobs — millions of them. The uncertainty surrounds which jobs will be lost, and what kinds of jobs will arise to replace them. Whatever else they achieved with their recently ended strike, Hollywood writers deserve credit for raising these and other important questions about how AI should be used.
One of the reasons for the strike was that writers are afraid their jobs will disappear and their skills devalued. It is a reasonable fear, given the rise of large-language models and Hollywood's reliance on predictable story lines. One of the sticking points in contract negotiations was how studios would use human writers and how they would credit generative AI.
What the writers got was minimum staffing agreements — for humans. They also got a promise that generative AI will not be credited as an author. And studios cannot use AI to write or rewrite literary material. At the same time, authors can choose to use AI as a tool.
Will this be enough to protect their jobs as technology evolves? Over the next few years, certainly. And that might be enough time — if they use it effectively to harness AI to be more productive, and if they allow their jobs to adjust.
Those are two big ifs. Understanding their importance is crucial not just for Hollywood writers, but for any worker in a changing labor market.
The US (and the world) has more jobs today than it did at the start of the 21st century. At the same time, plenty of occupations have declined due to technology. In 2023, there were only 32,000 people working in word processing and typing occupations, a sharp fall from 282,000 in 2000. Similar trends were seen in larger occupations such as sales and office workers, a category that has shed 6 million jobs since the start of the century.
Jobs typically evolve rather than disappearing outright. Some tasks are eliminated, some are added, until eventually the new version of the job no longer looks like the old one. Strictly speaking, the old job has “disappeared,” but if this evolution works well, the worker has not. For example, some office workers are now classified as managers because they have learned to use sophisticated software to manage human resources or payroll functions. So while there may be fewer office workers, there are more people in management positions.
In the past, technological change tended to reward the most skilled workers, because the technology was a complement to their skills. But generative AI can do things now done by the most highly educated workers; it can perform mental and creative tasks instead of physical or routine ones. So it's understandable that writers, lawyers, financial analysts, and the like are anxious about it.
A recent survey asked a panel of economists whether they thought AI would have a negative impact on the earnings potential of highly skilled workers. They were largely split: Most were uncertain, and almost as many agreed that it would. No one strongly disagreed.
And it's not just anxiety about jobs; there is also a lot of worry about education. Is a college degree still worth it? Here the surveyed economists were divided — a little more than half agreed that AI would lead to substantially greater uncertainty about the likely returns to investment in education, while nearly 40% disagreed.
My view is that the fear over declining returns to education is overblown. In 1980, college grads earned only about 40% more than those without a college degree. Today they earn roughly 80% more.
In theory, new technology allows humans to consume more while working less. In reality, however, technology displaces workers, as fewer people are needed to produce the same amount of goods or services.
Which brings us back to the Hollywood writers. Once armed with ChatGPT or other technology, they will need fewer hours to write the same quantity and quality of scripts. There is already research showing that writing quality and speed increases when people are trained to use ChatGPT effectively.
I count myself among those economists who believe generative AI will lead to rising returns to human skill. The most highly skilled people tend to be better positioned to adapt as jobs change. They often find ways to ensure that their skills are enhanced by technology rather than replaced by it.
To be certain, there are big unknowns. Universities need to make sure that students get skills that AI can complement rather than replace. Government needs to regulate AI to put appropriate guardrails around its misuse. There are difficult questions about ownership of intellectual property, and who is responsible and gets credit for writing. The writers agreement is a good start at addressing these issues, but ultimately national or even international standards will be needed.
AI creates a lot of risks. Thankfully, unions are leading the discussion about ways to address them. Now governments need to follow suit.
Elsewhere in Bloomberg Opinion:
- Unions Unite Against a Shared Enemy — AI: Brooke Sample
- AI Shines a Spotlight on Hollywood Hypocrisy: Parmy Olson
- Regulating AI Will Be Essential — and Complicated: Noah Feldman
For more Bloomberg Opinion, subscribe to our newsletter .
This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.
Betsey Stevenson is a professor of public policy and economics at the University of Michigan. She was on the president's Council of Economic Advisers and was chief economist at the US Department of Labor.
One more thing! HT Tech is now on WhatsApp Channels! Follow us by clicking the link so you never miss any update from the world of technology. Click here to join now!