The real choice confronting developing countries

Ravi Venkatesan

BENGALURU—In a recent commentary, Dani Rodrik, one of the world’s leading experts on the economics of trade and development, explains why he became a “manufacturing skeptic” after decades of seeing industrial production as the key to unlocking economic growth. Since traditional manufacturing cannot absorb the 1.5 billion workers “in occupations that neither require university education nor are exposed to the international economy through trade or offshoring,” services-led growth models must take center stage. Expand the opportunities in retail, hospitality, and food service, and middle-class consumption will drive productivity gains.

But largely missing from this analysis is the potential for AI-augmented gains in human productivity. Look around the world and you will find AI already quietly increasing productivity in labor-absorbing services in ways that neither require nor assume university education. This trend is especially apparent in India, which offers a path toward reconciling “manufacturing skepticism” with African and South/Southeast Asian policymakers’ persistent desire to industrialize.

India’s services-driven growth reflects more than an expansion of its retail and hospitality sectors. Algorithmic tools have fundamentally reshaped operations through inventory optimization, dynamic pricing, demand forecasting, and supply-chain coordination. As a result, a formal retail operation in Bangalore or Mumbai today is fundamentally different from a 1990s shop, not because the employees are more educated, but because algorithms have made them more productive.

The same dynamic can be found even at the level of street vendors and other micro-enterprises. With platforms like Flipkart B2B and JioMart offering informal merchants AI-powered demand forecasting and procurement optimization tools, a small vegetable vendor with only a mobile device can anticipate what customers will want, stock accordingly, and capture margins that were unavailable just five years ago.

Consider a parallel to another moment in industrial history. In the 1960s, a textile worker in South Korea produced probably 50 times more cloth than his Indian counterpart, not because he was any more capable, but because the loom was better. Now, AI does the same for street vendors, retail workers, and smallholder farmers. The big difference is that such tools can be made far more accessible. They are much easier to deploy than factories, and they work with existing informal structures, rather than requiring a wholesale industrial reorganization.

One implication of this difference is that it may not matter that 1.5 billion workers will remain in non-traded, low-skill occupations. Even if broad occupational patterns persist, productivity can grow. We already know that a smallholder farmer equipped with a digital agronomic adviser in East Africa or South Asia can realize average productivity gains of 30% without additional land or formal education. A vegetable vendor using simple AI tools for inventory and price optimization can double her earnings in 18 months.

The main constraint on productivity growth is not technological but institutional. Getting new tools into workers’ hands requires localization in native languages, accessibility without advanced digital literacy, business models that can be sustained with micro-margins, and integration with informal financial systems.

Crucially, policymakers must also ensure that these systems are built on inclusive open-access architecture. If dominant e-commerce platforms use predatory fees to extract the surplus value generated by inventory optimization, the street vendor’s hard-won margins will simply shift to corporate balance sheets. Checking all these boxes may not be easy, but it is feasible.

Similar findings apply to manufacturing. Although competing for a spot in modern global manufacturing value chains requires sophisticated skills, AI is fundamentally altering the equation. A garment factory no longer needs 500 multi-skilled workers; it needs 50 workers augmented by computer-vision quality control, demand-driven production planning, and logistics-optimization systems. One can already find such operations in Bangladesh and Vietnam.

To be sure, skeptics will point to the immediate displacement of the remaining 450 workers. But one must not overlook the broader economic transformation that will have been set in motion. As hyper-productive factories expand their output and lower unit costs, they catalyze a vast ecosystem of non-factory jobs—from upstream supply-chain coordination to downstream distribution—that can re-absorb labor at scale.

The productivity paradox that Rodrik identifies—the fact that expanded manufacturing employment no longer goes hand in hand with productivity growth—is real but not insoluble. When small-scale, informal manufacturers gain access to AI-enabled quality control, supply-chain coordination, and just-in-time planning, they can reach productivity levels formerly available only to large integrated factories.

Mexico’s weak economic performance under NAFTA supports this argument. The problem is not that Mexican manufacturers could not compete, but that technological diffusion failed to reach small-scale and informal operations. Factories expanded, but without the right tools and platform technologies, workers remained trapped doing low-margin assembly.

The same technologies can lower the human-capital requirements in many other sectors. While it is true that India owes much of its success in services-led growth to its large English-speaking population and early IT infrastructure investments, similar models can be replicated elsewhere. Thanks to advances in AI translation, you do not need English-speaking engineers to deploy digital agriculture tools where people speak Amharic, Swahili, or Telugu.

Thus, whether services-led growth is more feasible than industrialization in poor countries is beside the point. Feasibility depends only on the availability of productivity-enhancing mechanisms. In the 20th century, that meant access to industrial machinery. But today, it means access to AI tools that operate in local languages, work offline, and cost pennies per transaction.

Widespread use of such tools will not happen automatically, of course. Most AI investment today is concentrated in high-skill, high-wage economies. But this reflects policy choices, not some economic law. The question for policymakers is whether they will treat localized, affordable AI for informal workers as infrastructure—like rural electrification or highway systems—or as a luxury good for the wealthy.

If developing countries invest in making AI and related digital tools as widely available as possible, manufacturing- and services-led growth can coexist. A farmer using AI to improve yields can achieve a middle-class income without leaving agriculture, and a street vendor can do the same without joining a large, organized firm.

Manufacturing need not be the growth engine; it simply needs to coexist with AI-augmented formal and informal services. This vision may be less romantic than the East Asian industrialization narrative, but it has the advantage of being achievable right now. The main challenge is to implement a deliberate, inclusive AI-enabling strategy of infrastructure investment. The best way to lift up those 1.5 billion people is to meet them where they are.

Ravi Venkatesan, chairman of the Global Energy Alliance for People and Planet, serves on the boards of Hitachi and ServiceNow. This article was distributed by Project Syndicate.