Recent advances in artificial intelligence are spurring tectonic shifts across industries, and the pace of innovation will likely accelerate. According to the World Economic Forum’s Future of Jobs Survey 2024, 86% of employers anticipate that AI will drive business transformation in the next five years.¹ These technological advancements hold great potential for societal progress, fostering the discovery of new antibiotics that combat drug-resistant bacteria², for example, and the enhancement of agricultural yields.³ With this progress, however, comes risks: Who will gain? Who will lose? And how quickly will this revolution reshape the workforce?

The WEF survey demonstrates how AI may fit into an evolving skills marketplace, with employers saying about 40% of the “core skills” they demand will change by 2030. Additionally, 45% of employers surveyed deem AI and big data skills to be “core,” and nearly 9 of 10 report that these skills’ importance is on the rise. The AI transformation is largely yet to happen, meaning there is outsized uncertainty regarding AI’s actual impacts on the skills gap and the nature of work. Recognizing and describing these different forms of uncertainty will help us develop strategies for navigating the evolving AI landscape and mitigating potential risks to workers, firms and the nation’s economy.

What Labor Market Skills Can AI Replicate Now?

AI now excels at tasks like analyzing unstructured data, extracting key information from extensive documents, translating languages and generating code. Its capabilities, however, have limitations – notably, the well-documented tendency to produce inaccurate or fabricated information, often referred to as “hallucinations.” Building large language models and other AI tools that meet rigorous benchmarks for accuracy is the next frontier many firms are working to make a reality.

Thus far, AI investment has translated to a relatively modest impact on job demand in most industries. Yet in two professions – computer programming and translation – AI now seems poised to replace employees because, in these are areas, AI technologies have already demonstrated marketable capabilities.

Software engineering experienced a dip in demand around 2003, according to Live Data Technologies⁴, but it has since rebounded and is projected to grow by 17% from 2023 to 2033, as reported by the Bureau of Labor Statistics.⁵ While the bureau projects slower growth for translators (about 2% over the next decade), it does not forecast meaningful job cuts.⁶ Is the bureau exhibiting an AI blind spot? Or will AI’s impacts on labor be slower and more complementary than some are predicting?

For now, we should not expect AI to have caused widespread job losses. To significantly reduce employment for a specific role, the technological innovation would need to replace or make redundant nearly every task a worker performs. This total displacement is unlikely to happen soon, or at all. Rather than complete automation, AI is more likely to take over routine tasks in various occupations, supplementing instead of supplanting jobs. AI-powered diagnostic tools, for instance, can assist doctors, but they do not replace the need for human expertise. A recent study by OpenAI and academic researchers supports the notion that a sizable majority of American workers will find a helpful assistant in AI, not a potential replacement.⁷ The researchers estimate that 80% of US workers will have at least 10% of their tasks affected by AI, while 19% of workers are expected to see half or more of their tasks automated.

How Will Firms Deploy AI?

The speed of employer adoption is an under-discussed factor determining AI’s impact on jobs. Just because a technology can replace workers does not mean that it will. Cost-effectiveness is key.

Even if a technology offers cost savings, immediate uptake by firms is not guaranteed. Economic research finds that “frictions” often slow or prevent the adoption of cost-saving technologies. The upfront cost of adopting new technologies is a common friction that delays or precludes uptake despite potential long-term operating cost reductions. High interest rates exacerbate this issue, making it more expensive for firms to finance the initial investment.

Other factors beyond capital constraints also hinder technology adoption. For firms to embrace an innovation, they need to be aware of the technology and possess the skills to implement it. Some CEOs may be reluctant to take actions that would lead to layoffs, especially long-tenured executives who have strong connections with their employees.⁸ This reluctance is particularly pronounced during recessions, when layoffs are most detrimental to workers.

Because of these factors, AI’s impact will vary widely among firms and industries. Large firms with many roles to automate and capital to invest will be the earliest adopters, implementing new technologies to maximize their cost savings. Meanwhile, companies in tech hubs with a high concentration of talent will leverage advantages stemming from a readily available, skilled workforce needed to implement the new technologies.⁹

Share of Firms That Have Adopted AI by Industry

What Are the Implications for Skills Gaps in the US?

AI is likely to complement workers, yet its applications in the workplace will undoubtedly widen existing skill gaps while creating new skill vacancies. These technologies will automate certain tasks, reducing the demand for some skills that are now valued in the labor market. Yet the innovations will also enhance employee productivity by automating some redundant tasks and providing valuable insights. Increased productivity will likely drive demand for more workers, especially those with skills to effectively leverage AI tools and technologies.

Using the introduction of the personal computer and the internet as a guiding analogue, we should expect to see both trends – automation-induced redundancies and boosted productivity – take hold as AI adoption increases. The introduction of the personal computer most notably displaced workers performing routine tasks.¹⁰ Workers engaged in routine clerical work, for instance, suffered a devaluation of their skill set. At the same time, new skills in computing gained value and many new jobs were created. These new skills and novel positions, however, most often benefited workers different from those who were displaced.

The changing landscape of skill requirements is not a new phenomenon. Researchers estimate that more than 60% of jobs in the US are in occupations that did not exist in 1940.¹¹ As AI continues to develop, the needs of firms will evolve, creating new opportunities while disrupting existing roles, companies and industries.

A key feature distinguishing the present-day market from economies of the past is the unprecedented pace of change. Rapid technological development could disrupt labor markets faster than workers, businesses and governments can adapt. Yet the culprit, AI, could be a useful tool for mitigating these disruptions. By leveraging AI technologies to manage their own impacts, we could effectively retrain workers to acquire the new skills demanded by an AI-powered workplace. To achieve this efficiently, we must prioritize two strategies:

• Monitoring Labor Market Shifts: Continuously tracking and analyzing changes in labor market demands as AI technologies evolve.
• Developing Innovative Educational Programs: Creating new and accessible educational opportunities that equip current and future workers with the AI-related skills they need to succeed.

Why the Time May Be Right for AI

The future of AI’s capabilities is an area of high uncertainty, with lots of “known” and “unknown unknowns.” The innovation’s trajectory will be determined by technological advancements and evolving market needs. As the US economy faces the demographic headwinds of an aging labor pool and potentially reduced migration, the tightening labor market will likely increase the need for innovation. With fewer Americans of working age, we may need more machines.

Ultimately, while AI’s development and deployment present significant challenges, the technological innovation has the potential to lift firms and workers alike. If we are proactive in addressing risks, both realized and potential, and attentively manage AI-human interfaces, we can build a future in which AI’s benefits accrue across the economy.  

---

---

• Keyword(s):
• artificial intelligence 9
• innovation 188
• skills gap 13
• technology 80