Thinking through recombinant capital

By Jam Magdaleno and Cesar Ilao III

THE innovation race is leaving us behind. Even in the Global Innovation Index, where the country has inched upward in recent years, we are still being outpaced by most of our ASEAN neighbors. In AI Readiness, the gap is clearer: we remain in the bottom half of major Southeast Asian economies, trailing Singapore, Malaysia, Thailand, Indonesia, and notably Vietnam, which posts a score of 61.4 compared with the Philippines’ 58.5.

Why the performance? We do not lack talent in the Philippines. On the contrary, our creatives, BPOs, scientists, and engineers fare well globally.

Our view of innovation under the Fourth Industrial Revolution rests on two serious errors: the vanity of trying to be a “pioneer innovator,” essentially the next Silicon Valley, and that the government must be the driver of this innovation.

The first error is simply unrealistic. It’s too late to compete with nations like China or Singapore for top-innovator status — both technologically decades ahead. Such a goal leaves us in a perpetual state of anxiety that our industries, particularly BPOs, are one step away from obsolescence.

The second error is naive. The government is structurally inefficient as an engine of innovation. It always faces the gargantuan challenge posed by the “knowledge problem” —because the state enjoys a monopoly on decision-making, it engages in the serial processing of knowledge, and hence can only conduct one or few singular trials at a time. This creates a crisis of counterfactual evaluation. Unlike a scientist, the government has no control group for comparison. In contrast, markets operate through parallel processing, where the simultaneous experiments of competing firms function like trials in a laboratory. By relying on the state’s singular approach rather than the market’s parallel discoveries, we inevitably stifle the very innovation we seek.

A good example is the failed Jeepney Modernization Program. By mandating a singular, top-down template, the state forces the entire country into one “serial trial.” This is like a giant science experiment where the lead scientist refuses to have a control group. Because the government has a monopoly on the rules, we are all stuck on one path. We have no way of knowing if a different approach (like app-based dispatching, locally manufactured models, or engine retrofitting) would have been faster or cheaper, because the government never allowed those “parallel experiments” to happen. While the market works like a laboratory where hundreds of firms test different ideas at once, the state’s “singularity” forces us to wait for one slow, potentially flawed experiment to finish before we can try anything else.

Some might propose a Chinese-styled “engineering state” as the solution, as described by technology analyst Dan Wang. But this presumes an efficient developmental apparatus, a premise that collapses in our context. Here, the theoretical cost of state-led projects is frequently bloated by corruption spillovers, with leakage reaching staggering levels of 25% to 70%, as we have seen in the recent flood control controversies. This makes state-directed innovation not just inefficient, but fiscally predatory.

Instead of these assumptions. We must shift our view of innovation from the “invention of the new” to the “recombination of the existing.” As Harvard economist Martin Weitzman argued, long-run growth does not arise from simply adding more capital and labor — a process subject to diminishing returns, but from recombining existing ideas and assets into new, more valuable combinations. In this sense, the “recipe” matters far more than the quantity of ingredients.

The Philippine economy’s unique composition is better suited to recombination, leveraging existing capital and applying global technologies, rather than attempting to invent them from scratch. We hold a distinct advantage in sectors such as AI and the creative industries, underpinned by an asset that requires no government procurement to mobilize: a large, adaptable, English-speaking workforce.

We do not need to invent the next OpenAI, which is a game of massive capital. We need to master the art of splicing our existing, undervalued assets with emerging technologies to create unique value propositions.

On AI, our response remains on the defensive. We view it as a hurricane threatening to wipe out our BPO sector, prompting calls for “upskilling.” But this is a deficient solution. A recombinant approach sees the BPO sector with realism. While the threat of automation exists, we possess two massive, dormant assets: a workforce of 1.2 million culturally adaptable English speakers, and decades of unstructured data on Western consumer psychology (contained in call logs and chat histories).

While Singapore invests billions in “Compute” (hardware and data centers), the Philippines could invest in “Context.” The recombinant move is to merge our service workforce with Large Language Models (LLMs). We should not just be users of AI. We should be the global hub for vertical AI alignment. Generic models like GPT-4 are powerful but lack nuance. The Philippines can recombine its workforce’s “cultural empathy” with algorithmic power to create specialized, industry-specific models. For example, instead of just answering calls, Filipino teams could be training “empathy engines,” fine-tuning AI agents for US mental health or triage or geriatric care, where nuance is non-negotiable.

This shifts the value proposition from selling time (labor) to selling high-fidelity context (intellectual property). This requires zero investments on new bridges or airports, but relies on a regulatory environment that allows data and talent to mix freely.

Our fallacious view of innovation renders our approach reactive, even as our regional neighbors play a different game. Vietnam’s Green and Smart Mobility (GSM), for instance, didn’t wait for a state-mandated roadmap, nor did it invent EV. It succeeded by recombining existing capital — VinFast’s electric vehicle manufacturing with a high-tech digital ride-hailing platform. By merging hardware with software, GSM captured nearly half of Vietnam’s ride-hailing market in just two years and is now exporting its model across Southeast Asia, including the Philippines. Vietnam’s government has supportive policy environments for EV adoption, yet GSM’s expansion and market share are driven by corporate strategy, not state direction.

The government can also learn from the boy band SB19. Why? They did not succeed by simply mimicking OPM or K-Pop. They recombined existing capitals: the rigorous, algorithmic training capability of the Korean idol system and Filipino linguistic depth and vocal talent. They treated “performance discipline” as a capital asset and merged it with local creativity. No wonder they command global attention today.

SB19 learned to stop treating Filipino culture as a museum artifact and realized that, like any form of capital, it can be remixed into something new.

Why does this matter? Because this shift in perspective brings us back to the fundamental building blocks of innovation: property rights, enforceable contracts, and the rule of law. At a minimum, the government’s role is to ensure these elements are consistently observed and protected. Can entrepreneurs get permits without bribery? Can they access financing and technology? Can they hire talent and import materials freely? Can they experiment without the threat of monopolies?

If the government can’t get the basics right amid recent scandals, how can it claim to be the “driver” of innovation?

Jam Magdaleno is head of Information and Communications at the Foundation for Economic Freedom (FEF) and an Asia freedom fellow at the London School of Economics and Political Science (LSE) and King’s College London. Cesar Ilao III is a researcher and communications specialist for FEF. He is a lecturer at the University of the Philippines and was formerly a researcher at Monash University, Australia.