AI is no longer in its infancy, but much of its economic impact is yet to be realised.
That’s a key takeaway from a new report from the McKinsey Global Institute, which argues AI could potentially deliver additional economic output of around $13 trillion by 2030, boosting global GDP by about 1.2 per cent a year.
The report, Notes from the frontier: Modeling the impact of AI on the world economy, models trends in AI adoption and explores how the family of technologies may kick off a competitive race with major implications firms, labour markets, and economies more broadly.
“The productivity dividend of AI probably will not materialise immediately. Its impact is likely to build up at an accelerated pace over time; therefore, the benefits of initial investment might not be visible in the short term,” the authors write.
The researchers argue AI adoption may follow an S-curve pattern, “a slow start given the investment associated with learning and deploying the technology, and then acceleration driven by competition and improvements in complementary capabilities.
“As a result, AI’s contribution to growth may be three or more times higher by 2030 than it is over the next five years. Initial investment, ongoing refinement of techniques and applications, and significant transition costs might limit adoption by smaller firms.”
The McKinsey Global Institute looked at five broad categories of AI: computer vision, natural language, virtual assistants, robotic process automation, and advanced machine learning. Its modelling expects 70 per cent of firms of adopt at least one form of AI by 2030.
However the economic benefits will not be evenly distributed.
The report argues front-runners could potentially double their returns by 2030, while laggards may experience around a 20 per cent decline in their cash flow from today’s levels.
The AI leaders, defined as companies that fully absorb AI tools across their enterprises over the next five to seven years, tend to have a strong starting base for IT, a higher propensity to invest in AI and positive views of the business case for AI.
While the laggards, companies that do not adopt AI technologies at all or have not fully absorbed them in their enterprises by 2030, will find it difficult to catch up.
The report identified five factors that are currently limiting the applications of AI in business.
- The human capital required to label training data. “In supervised learning, machines do not learn by themselves but need to be taught, which means that humans must label and categorise the underlying training data.”
- Accessing large enough data sets to train algorithms. “At present, the availability of labeled data is critical since most current AI models are trained through supervised learning, and categorising data correctly requires a huge amount of human time.”
- Difficulty explaining the results from large complex neural-network based systems. “One development—still at an early stage—that could improve the ease of explaining or transparency of models is local interpretable model agnostic explanations, which attempt to identify which parts of input data a trained model relies on most to make predictions.”
- Difficulty generalising. “AI models still have difficulty carrying their experiences from one set of circumstances to another, which leaves companies having to commit resources to training new models even if use cases are relatively similar to previous ones.”
- Risk of bias. Unlike the first four, this issue can’t be solved by technology. “A great deal of academic, nonprofit, and private-sector research is now underway on this issue.”
The research also examines the economic impact of AI on countries and individuals. Nations that lead the adoption of AI could capture 20 to 25 per cent more in economic benefits compared to current levels, according to the report.
“Policy makers will need to show bold leadership to overcome understandable discomfort among citizens about the perceived threat to their jobs as automation takes hold,” the authors write.
“Companies will also be important actors in searching for solutions on the mammoth task of skilling and reskilling people to work with AI.”