FIGURE 3.9 AI or ML Software Is Used More Widely in High-Income Countries Number of firms that purchased AI or ML software, top 15 countries, 2018 United States United Kingdom India Canada France Germany Australia Netherlands Spain Italy Switzerland China Singapore Brazil Ireland 0
2,000
4,000
6,000
8,000
10,000
12,000
Number of firms Source: Calculations based on iDatalabs (now Enlyft, http://www.enlyft.com) data for the World Bank. Note: The dataset is generated using web scraping. It covers 17,000 formal firms in 107 countries across regions and income levels. AI = artificial intelligence; ML = machine learning.
skills, changes in organizational design and business models, legal constraints, and even cultural expectations (Brynjolfsson and Mitchell 2017).
The Rise of Intangible Capital Chapter 2 showed that except in transportation, warehousing, and telecommunications, physical capital has typically played a much smaller role in the production process among services firms than among manufacturing firms. The physical capital per worker for low-skill services sectors such as retail or accommodation and food services is a third of that in manufacturing, and it was also lower in many global innovator services such as programming and IT. The question is the extent to which the steady rise of firms’ investment in intangible capital over the past 20 years—often even exceeding growth in tangible capital (Corrado et al. 2018)—will affect the services sector. Evidence from high-income economies suggests that for every £1 of investment in tangible capital such as buildings or machinery, firms spent £1.10 on intangible capital in 2013 (Haskel and Westlake 2018). Intangible capital can be classified into three broad categories: (a) computer-related software and data; (b) properties of innovation such as R&D, design, and artistic originals; and (c) company competencies such as marketing and branding, 146
At Your Service? The Promise of Services-Led Development