Posted at AEI Ideas

From Automation to Reinvention: How AI Is Shifting the Nature of Work

The latest breakthroughs in artificial intelligence (AI) will bring numerous productivity improvements to businesses but are likely to change the nature of work itself, which could disrupt millions of workers. 

Three factors are driving the rapid adoption of AI technologies. First, the pandemic compelled companies to undergo digital transformations. A McKinsey survey found that over 67 percent of CEOs accelerated AI and automation adoption during this time.

Second, persistent labor shortages are causing companies to adopt automation instead of hiring more workers. For example, a shortage of truck drivers is driving the development of autonomous vehicles for use in construction, farming, and heavy industry. North Dakota’s Gov. Doug Burgum has even proposed a $5 million expansion of the state’s automation tax credit to help businesses adopt new technologies.

Finally, the emergence of large language models (LLMs) such as OpenAI’s ChatGPT, Microsoft’s Bing Chat, Google Bard, Anthropic‘s Claude, and MidJourney has not only raised awareness of AI capabilities but also led to a growing number of new productivity tools.

It’s been four months since ChatGPT first captured the country’s imagination and we’re beginning to see the first set of economic impact estimates. Goldman Sachs predicts a 7 percent global GDP boost from generative AI over a decade but warns of significant labor market disruption, with 63 percent of US jobs potentially affected by AI automation.

OpenAI and the University of Pennsylvania released a working paper estimating that 80 percent of the workforce could see at least 10 percent of their tasks affected by large language models. And for 19 percent of jobs, nearly 50 percent of their functions could be “exposed” to these AI tools.

One shared observation in both reports is that in a departure from prior automation waves, higher-skilled and higher-income professions stand to be most affected. AI technologies possess the capacity to undertake at least half of the tasks currently performed by professionals such as auditors, interpreters, and writers.

It is also important to distinguish between jobs “exposed” to AI and those displaced by it. Workers will increasingly interact with AI technologies, which will alleviate the burden of mundane, tedious tasks, enabling them instead to focus on more advanced and valuable work. Box CEO Aaron Levie aptly expressed this notion, stating, “The impact of AI for the vast majority of work will be as an accelerant to getting to the next most important task that only a human can do.”

These tools are also likely to lower the barriers for workers to enter new jobs or perform higher-skilled work. Two graduate students from the Massachusetts Institute of Technology’s economics department conducted a study with professionals in fields such as marketing and human resources. Half of the participants incorporated ChatGPT into their daily tasks while the other half did not. As expected, ChatGPT increased overall productivity. However, the most intriguing finding was that ChatGPT significantly benefited the least skilled and accomplished workers, narrowing the performance gap among employees. In essence, weaker writers got much better while good writers simply became more efficient.

What does this all mean? It’s still too early in this wave of AI technologies to fully know, but it seems reasonable to assume that several productivity tools will be quickly deployed across a wide range of occupations. That means that every new job will become slightly higher-skilled as workers need to learn to use these new AI systems for their specific tasks. 

While these technologies will displace some jobs, it seems more likely that the biggest disruption will be the new jobs these technologies create. History offers important lessons from when other technologies were introduced to the economy. For example, the adoption of spreadsheets led to fewer bookkeepers but fueled a surge in financial managers and accountants. AI is likely to create similar trends, including the introduction of new occupations, such as prompt engineers, or create new hybrid jobs that combine domain expertise with new AI technical skills.

Policymakers can take proactive steps to smooth out these transitions’ volatility. As the US Chamber of Commerce’s Commission on AI Competitiveness, Inclusion, and Innovation recommended, the United States must increase education around AI in K–12 and higher education systems. Efforts such as aiEDU and Code.org will be important in the years ahead. And public and private sectors need to focus on training and reskilling the workforce through targeted investments in programs that facilitate worker transitions and provide incentives for businesses to invest in retraining.

The rapid adoption of AI technologies promises significant productivity gains and economic growth, but it also necessitates a workforce reevaluation as new AI systems reshape the landscape of higher-skilled professions, potentially altering the nature of work and driving the need for adaptation and reskilling in the face of unprecedented change.