Page 102 - ISC PROCEEDINGS 21.4

More recent theoretical work extends the Solow framework to incorporate AI as an
endogenous growth driver. Drawing on Romer's (1990) endogenous innovation model,
Ijmra (2026) argues that AI creates recursive feedback loops between technology,
knowledge, and output that redefine long-term growth trajectories, potentially
overcoming the diminishing returns of the classical Solow model. This AI-augmented
growth perspective suggests that economies investing early and systemically in AI
infrastructure may achieve compounded TFP advantages relative to later adopters, a first-
mover premium with direct implications for Vietnam's industrial strategy.
Figure 1 illustrates this system: foundational inputs on the left (human capital, data
infrastructure, institutional frameworks including the AI Law, dedicated funds, and special
economic zones) feed into AI systems (models, platforms, semiconductors) at the center,
which generate substantial value only when paired with complementary organizational
changes, workflow redesign, skills upgrading, and new business models. The reinforcing
feedback between AI capability and organizational transformation ultimately produces
firm-level outcomes (higher TFP, cost reduction, quality gains), sectoral transformation
across industry, healthcare, agriculture, and digital government, and macro-level gains in
GDP growth, labor productivity, and green development, with the J-shaped curve
underscoring the delayed but ultimately accelerating nature of AI's productivity impact.

Figure 1. Inputs and outputs of AI applications
Source. Compiled by the authors based on the GPT framework and Vietnam’s policy
documents
2.4. Literature review of AI, productivity, and digital economy in Vietnam
The academic literature on AI and economic productivity in developing Asia is
growing but uneven. At the global level, McKinsey Global Institute (2023) estimated that
generative AI could add $2.6-$4.4 trillion annually to the global economy, with labor
productivity growing 0.1-0.6% per year through 2040. The UNDP projects AI could add
approximately 2 percentage points to Asia-Pacific GDP annually, while the OECD forecasts
AI's contribution to the world economy could exceed $15 trillion by 2030 (OECD, 2025).
Vietnam-specific research has expanded substantially since 2020. Nguyen (2024) finds a
positive but qualified relationship between digital transformation and economic growth,
contingent on IT application quality and human resource infrastructure. Tran and Le (2025)
document a declining trend in science and technology's contribution to TFP in Ha Tinh
province, attributing it to technological saturation without corresponding organizational

101