]The discussion paper is published by McKinsey Global Institute (MGI), the business and economics research arm of McKinsey & Company, in September 2018.

This paper focuses on modelling AI’s potential impact on the economy and based on a micro-to-macro and simulation-based approach, it gives a balanced view on both the possible benefits and the costs associated with adoption of AI.

The point that the paper tries to highlight is that AI has the potential to deliver additional economic output of around $13 trillion or 16% to global output by 2030 thereby boosting global GDP by about 1.2 percent a year. An early average simulation indicates that by 2030, 70% of companies will adopt any one type of available AI technologies. However, the adoption of such technologies will typically follow a S-shaped curve justified by a slow start given the huge amount of initial investment followed by competition and improvements in complementary capabilities. Hence, realization of full economic impact of AI deployment can be gradually observed over a period of 5 years.

Despite the gains, the report clearly states that adoption of AI could widen the performance gap between countries, companies and workers with early adopters capturing an additional 20-25% economic benefits. Nevertheless, in the end, what matters the most is the extent to which countries or companies choose to embrace and deploy AI to realise the full economic impact.

Relevance of the Report

In the past few years, the use of AI tools and techniques have gained momentum across businesses and hence it becomes important to understand on the one hand AI’s potentiality in contributing towards global economic activity while on the other the disproportionate benefit that are likely to happen for front-runners and laggards. Such imbalanced growth can mostly be attributed to strong starting digital base or simply a higher propensity to invest in AI-related technologies. In this regard, it can be pointed out that this paper deeply analysed seven possible channels for AI impact. They are augmentation, substitution, product and service innovation and extension, economic gains from increased global flows, wealth creation and reinvestment, transition and implementation cost and negative externalities. While the first three factors relate to the impact of AI adoption on the need for production factors that have direct correlation with firm productivity, the remaining four factors are externalities linked to such adoption. However, the interesting thing is that only 3 channels i.e., 1) the use of AI-driven automation to substitute existing labour; (2) the application of AI to innovation that creates new and better products and services; and (3) AI-driven competition and the resulting disruption to firms and workers, stand out as they have the potential to add $6-$9 trillion to global GDP by 2030. However, on the down side, negative externalities could reduce the gross GDP impact by $7 trillion.

Unlike other studies where positive impacts of AI is emphasized, this report helps readers to form an unbiased view about the adoption and absorption of AI as both micro and macro factors contribute to the impact of AI.

Key Takeaways

• Despite AI’s tremendous potential to contribute to global GDP, it can widen gaps between countries, organizations and workers. With explosion of data, improvements in computing power and capacity and progress in algorithms, AI’s reach has widened across industries.

• The impact of AI can be felt over time; gathering pace after five to ten years. While the small initial impact can lead to hyped judgement about adoption of AI, the benefits to early adopters becomes visible only in later years.

• Early digitalization and the urge to stay ahead in the competitive race are some of the key determinants for the pace of AI adoption. Other factors include AI investments and related R&D, human capital, labour market structure and flexibility. 

• With varying degree of AI adoption across countries, there will be global leaders and countries with moderate to weak foundations. The more the absorption rate diverge, the more diverse will be the economic impact. Countries with capability to innovate could generate 10% of such impact. Similarly, front-runner companies could increase economic value by 122%.

• As large-scale adoption of AI is likely to disrupt labour market, there will be an increase in demand for certain skills thereby widening the wage gap. Despite fears that automation will wipe off jobs, AI will have neutral to modest negative impact on long-term overall employment.