A new study analyzes the potential of generative AI as a general-purpose technology. According to the study, AI could boost productivity, accelerate growth, and change labor markets faster than previous foundational technologies - posing challenges even for highly skilled workers.
Generative AI has the potential to profoundly transform the economy and society as a so-called general-purpose technology. Like the steam engine, the internal combustion engine, and electrification before it, this new AI technology could significantly accelerate productivity growth. This is the conclusion reached by Andrew McAfee, researcher at the MIT Sloan School of Management and co-founder and co-director of MIT’s Initiative on the Digital Economy. The study was conducted as part of a Technology & Society Visiting Fellowship at Google.
McAfee argues that generative AI meets the three characteristics of a general-purpose technology: it is rapidly improving, widely applicable, and enables complementary innovations. Within a few years, language models have made tremendous performance leaps, such as in knowledge tests and legal exams. At the same time, studies show that the technology could already make at least 10 percent of tasks more productive in 80 percent of occupations. New applications, such as in robotics and materials research, hint at the innovation potential.
Unlike earlier general-purpose technology that have spread over decades, generative AI could have a much faster impact, according to McAfee. The necessary digital infrastructure is already in place, users don't need special computer skills, and productivity gains are visible after just a few weeks. Companies with AI-savvy employees have seen higher stock market growth than less exposed companies since 2022.
Goldman Sachs estimates that generative AI could accelerate economic growth in the US by 0.4 percentage points per year over the next decade and by 0.3 percentage points in other industrialized countries. This additional growth could lead to greater prosperity, better health, and less environmental pollution, McAfee argues. In addition, the technology could relieve people of routine tasks and give them more time for value-adding activities, such as in medicine.
Some highly educated workers will have to shift what they do
The AI boom also poses risks. As with previous upheavals, the demand for skills is likely to change dramatically. "Current technologies have demonstrated expert-level performance at tasks requiring creativity, analysis, problem solving, persuasion, summarization, and other advanced skills," says McAfee. "Some highly educated and well-paid workers who make a living via these skills today will have to shift what they do in order to maintain or increase their value in the labor market."
However, generative AI will not necessarily automate all tasks. According to McAfee, research results show that the best results are achieved when humans continuously interact with the technology. Mastering this interaction will be a valuable skill for people in many professions, even for those with the highest qualifications and education levels.
So far, historical experience does not point to technology-driven mass unemployment. But income inequality and disruptive structural change could increase as skills and jobs change rapidly, the study says. Another risk is that large companies could further expand their lead while smaller ones fall behind. To cushion the social impact of the transition, McAfee suggests using successful approaches from the pandemic, such as wage replacement benefits and retraining. AI systems themselves could also facilitate lifelong learning.
Which countries will be ahead in the AI age is an open question. Success factors are likely to include a market-oriented innovation culture, digital infrastructure, specialized education, and a stable legal framework. One thing is certain, according to McAfee: those who drive the development and use of generative AI and regulate it wisely can benefit economically. Countries far from the technological frontier, on the other hand, risk falling behind.
