We are living through humanity’s fourth industrial revolution, which is largely driven by breakthroughs in digital technologies. Some, like the Internet and artificial intelligence (AI), are converging and amplifying each other, with far-reaching consequences for economies and societies. For developing countries, the implications are profound, and questions concerning policy choices and the “appropriateness” of new technologies have become urgent.
Even if new technologies seem likely to fuel unemployment and deepen income inequality, no country can simply reject them outright. Instead, policymakers must understand the multifaceted and complex nature of a technology’s appropriateness (or inappropriateness) for development, and then pursue nuanced responses that aim to maximize the benefits and minimize the harms.
In development economics, an appropriate technology is defined as one tailored to fit the psychosocial and biophysical context prevailing in a particular location and period. Such tools are designed with a view to the environmental, ethical, cultural, social, political and economic aspects of the communities for which they are intended. A technology’s appropriateness for development thus can manifest across many dimensions.
Illustration: Louise Ting
For example, compared with technologies from Europe and the US, those from China and India tend to be more appropriate for the conditions prevailing in the least-developed countries. Technologies suited to sub-Saharan Africa, for example, include hand pumps, pharmaceuticals, mobile phones and solar energy. By contrast, automation technologies designed to address the needs of Japan’s aging society would not be appropriate for low-income countries with massive youth populations in need of work.
The current wave of emerging digital technologies can be grouped into three categories: efficiency-enhancing ones such as AI and robots; connectivity-enhancing ones such as Internet-connected devices (from mobile phones to the Internet of Things), digital platforms and virtual reality; and infrastructural ones such as 5G, cloud computing and big data.
Let us focus on the efficiency and connectivity-enhancing categories, which are the applied technologies directly used by organizations and individual users.
My own analysis of their appropriateness reveals a complex picture across many dimensions, including the economic, the technical, the social, the environmental, the ethical and the cultural. For example, in the cultural dimension, a technology’s appropriateness might hinge on a given society’s expectations of individual privacy. These can differ widely: The expectation of privacy online is significantly lower in China than in the EU (with its “right to be forgotten” law).
While efficiency-enhancing technologies promise increased productivity by reducing labor costs in production, we know that widespread adoption of industrial robots and AI could create serious social and economic challenges in terms of employment and income inequality. Though we have yet to see job replacement at scale, the potential is certainly there. Moreover, developed countries’ relocation of capabilities newly amenable to robotization threatens to close the window of opportunity for less-developed countries to pursue industrialization through manufacturing.
AI and industrial robots also require substantial data-storage capacity, processing power and analytical capabilities — a high entry threshold that would prevent developing countries from adopting them quickly and catching up. The large-scale deployment of AI would introduce many ethical challenges as well, meaning that the clock is ticking for policymakers to establish safeguards and other measures to minimize harm.
As for connectivity-enhancing technologies, the economic benefits come in the form of lower access costs and enhanced economies of scale. By lowering entry barriers, these technologies can help to include marginalized communities in value creation, as well as improve access to financial and educational resources and information, and health and other public services.
From the supply side, connectivity-enhancing technologies can create opportunities for widespread adoption by workers and consumers. Easier and more timely access to information can lead to entirely new models of value creation. While these tools require digital infrastructure and basic digital skills, the threshold is lower than it is for AI and big data.
Moreover, innovations like mobile Internet give developing countries the opportunity to leapfrog past traditional cabled communication technologies that were unavailable or too expensive and technically difficult to scale up. However, of course, these technologies also raise ethical challenges when it comes to cybersecurity, social stability, privacy, public trust and so forth.
New technologies always facilitate new ways of working and consuming. However, to map future trends in manufacturing, we should look to where the different categories of digital technologies interact and reinforce one another. How they diffuse and are adopted would define the next phase of technology-enabled productivity.
Two scenarios stand out. First, AI and industrial robots might soon become widely viable. If so, there would be more and faster, relocation of manufacturing to industrialized countries, as well as increased concentration of manufacturing in fewer large manufacturing hubs and countries. Manufacturing would remain an important driver of income growth and industrialization, but it would no longer be the primary engine of job creation. Income inequalities between countries would widen.
Second, few commentators have yet to grapple with the transformative and disruptive potential of 3D printing, which could replace the mass-production model of manufacturing. This technology — which is significantly enhanced by AI — has come a long way, and is now poised to replace the traditional assembly line with more decentralized and bespoke production systems located closer to the consumer. If current trends continue, we could see a dramatic compression of the global value chain into one machine.
Faced with these trends, policymakers in developing countries would need to focus on four priorities. The first is to accelerate digitalization. It has long been apparent that digital technologies have the potential to be as revolutionary as electricity was for a previous generation. The sooner developing countries embrace and adopt them, the better chance they would have to keep up, or even to leapfrog over incumbent tools and methods. Conversely, rejecting new technologies all but ensures that one would be left further behind.
Developing countries should do what they can to increase investments in digital infrastructure. That could mean installing broadband, extending 4G or even 5G to wider areas beyond major cities, building facilities for big data storage and analysis, developing digital skills across the labor force, helping small and medium-sized enterprises pursue their own digital transformations, building regulatory capacities to supervise digital development, and, if the opportunity presents itself, developing robots or AI-empowered production capacities in industries where the country does not have prior capacity.
Here, speed is of the essence, because AI, industrial robots and 3D printing have not yet become economically viable on a global scale. Until they do, the relocation of global production and value chains would proceed at a relatively slow pace, and developing countries would still have a (narrowing) window of opportunity to catch up through manufacturing-based industrialization. However, this would be a new model of industrialization, fundamentally different from the path previously taken by developed countries, and more recently by countries like Japan, South Korea and China.
The new industrialization necessarily would build on the current digital revolution. A coherent vision of the future of production and work would be incorporated in its design, as would various characteristics of each country’s particular development strategy.
The infrastructure requirements, production processes and business models would all be different from those used in traditional industrialization. To have a chance, developing countries must consider all of these variables when designing and developing their industrial bases and the capabilities that continued competitiveness requires.
A second priority is to seize on the development opportunities of a digitalized services sector. A substantial body of research shows that the mobile Internet, digital platforms, and the gig economy have had a significant impact on development, notwithstanding all the governance challenges that come with them. Countries that have embraced these innovations, especially within the infrastructure and connectivity-enhancing categories, have been able to expand the reach of existing services and products, as well as empower individual innovators, especially in marginalized communities. Technology-based entrepreneurial activities can completely transform a society as new entrants build on what others have created.
Moreover, efforts to integrate the digital economy with traditional manufacturing and services production would enhance those firms’ productivity and competitiveness, allowing them to expand their market.
Digital technologies also make some nontradable services tradable, by disembodying services from their providers, as we are already seeing in education, healthcare, mobile banking, and video streaming. These digitalized, more tradable services can then become an important driver of economic growth. True, there is still a debate about the extent to which a services-driven economic development model can work. Even so, the income growth, job creation, and welfare and empowerment benefits that come with digitalized services are important to any country.
A third big issue is the future of work. While manufacturing would remain an engine of income growth, it would be a far less powerful engine of job creation. The organization of employment in many sectors, especially knowledge-intensive services, would become more flexible and decentralized through remote work. Though more blue-collar workers would be phased out of large manufacturing workplaces, connectivity-enhancing technologies would help to reduce the barriers facing marginalized communities at the bottom of the pyramid.
Again, there is a debate over the desirability of these changes overall, because blue-collar work in large manufacturing companies has long been a source of financial stability and upward social mobility. Freelance and remote work would require more self-discipline, self-motivation and self-organization. Still, these workers also would enjoy greater autonomy, flexibility and a better work-life balance. With proper training and skills development, they might have more freedom to do what they prefer. If so, we would see widespread disruptive changes in the labor market, as well as new patterns of production, consumption, commuting and so forth.
Last but not least, policymakers need to stay focused on the potential dark side of emerging technologies. There has been much debate about introducing a “robot tax” to contain the negative effects of AI and industrial automation. However, even more important are policy incentives to guide the flow of human and financial resources from research and development to innovation and commercialization. For example, one idea is to introduce an appropriate technology score, which would then bear on decisions relating to R&D, technology transfers and investment decisions. This kind of ex ante approach would be more effective than ex post taxes on the deployment of AI and robots in production.
Some might object that incentivizing appropriate technology could cause firms, sectors or countries to fall behind the technological frontier. However, appropriate technology does not mean less-advanced technology. Mobile Internet is more appropriate for the developing countries that lack Internet cables and face economic, technical and geographic constraints in deploying such infrastructure. Likewise, small agricultural machines are easier to use in mountainous fields and solar panels are ideal for remote, desert or tropical regions.
Above all, government policies and international cooperative efforts should emphasize the development of infrastructure, skills and regulatory capacity in the digital economy.
These are the ingredients needed to ensure that developing countries can build up needed competences and compete in the current industrial revolution. The path that today’s industrialized countries took is no longer open. The digital age requires a new modernization model.
Fu Xiaolan is founding director of the Technology and Management Centre for Development, founder of OxValue.AI, and professor of technology and international development at the University of Oxford.
Copyright: Project Syndicate
RSS