Rethinking economy: Digital labor redraws line between capital and human input

As artificial intelligence systems continue to permeate workplaces and reshape productivity dynamics, a new economic framework is being proposed to account for their distinct contribution. A study titled “Evolving the Productivity Equation: Should Digital Labor Be Considered a New Factor of Production?” published on arXiv, argues for a conceptual shift in how AI is measured within productivity and growth models.

The paper challenges the current reliance on Total Factor Productivity (TFP) to account for AI-driven outputs, and recommends recognizing “digital labor” as a distinct factor of production, alongside capital and human labor.

Why traditional economic metrics fail to capture AI's true value

Current productivity metrics were built for a material economy, where outputs were tangible and directly tied to human labor or capital investment. In today’s digital economy, however, much of the value created, through software, data, and AI-driven services, remains unmeasured or misclassified. The study highlights that GDP calculations and firm-level KPIs often underrepresent contributions from generative AI and other cognitive systems because they are either categorized as operational expenses or bundled indistinctly into residual TFP.

This statistical blind spot has widened as AI begins to replace or augment cognitive functions across industries. For example, when AI systems streamline logistics, enhance code generation, or enable real-time customer service, their impact on productivity and quality is rarely captured in national accounts or firm balance sheets. The International Accounting Standards (IAS 38) framework, for instance, excludes many internally developed AI tools from being treated as capital, leading to further misalignment between economic value creation and reporting.

By remaining invisible within existing frameworks, AI contributions are obscured, limiting both policymakers' and business leaders’ ability to assess their strategic importance. The authors argue that continuing to absorb AI into TFP - a statistical “black box” - is increasingly insufficient for tracking the modern drivers of economic growth.

What makes digital labor a distinct economic input?

Digital labor, as defined by the study, refers to the autonomous cognitive capabilities of AI systems, ranging from chatbots and diagnostic algorithms to code-generating agents and creative assistants. Unlike traditional capital or human labor, digital labor possesses five defining characteristics: scalability, intangibility, self-improvement, volatility, and elastic substitutability with human roles.

Digital labor is infinitely scalable at near-zero marginal cost. Once an AI model is developed, it can be deployed across millions of users and tasks without degrading performance. This non-rivalrous yet excludable nature allows AI to produce compounding returns, which starkly contrasts with the limited scalability of both physical capital and human labor.

In addition, AI systems can self-improve by learning from data and feedback. Over time, this makes them more efficient and accurate - an attribute that parallels learning in human labor but occurs at vastly different speeds and cost structures. However, this self-improvement comes with rapid obsolescence. AI models can degrade quickly if trained on stale or recursive data, and new algorithmic breakthroughs can render previous generations economically obsolete overnight.

Furthermore, AI's substitutability with human labor is highly nuanced. In structured, rule-based environments, AI may act as a direct replacement. In creative or socially complex roles, it serves more as an augmentation tool. This creates a non-linear substitution dynamic, where AI might both replace and multiply human labor depending on the task and context.

Recognizing digital labor as an independent factor, rather than forcing it into legacy categories, allows organizations to model its impact more precisely. It also parallels historical shifts, such as the recognition of human capital in the 20th century, which improved our understanding of growth by quantifying the value of education and skills.

How should businesses strategically adapt to digital labor?

The integration of digital labor into classical growth models like Solow's and Romer's offers significant strategic implications for businesses. In the Solow framework, digital labor boosts productivity by enhancing the efficiency of existing capital and labor. In the Romer model, it accelerates knowledge generation and innovation. Both effects point to the need for deliberate investment in digital labor not just as a cost-saving tool, but as a growth multiplier.

To adapt, firms must recalibrate how they allocate resources. Leaders are encouraged to optimize the interplay between human labor, capital, and digital labor. This includes determining which tasks are best handled by AI, which require human expertise, and which benefit from hybrid collaboration. Routine, data-intensive processes can be offloaded to AI, while creative and ethical judgments remain human-led.

Importantly, the study urges companies to develop new performance metrics to track AI contributions separately. Rather than treating AI outputs as indirect improvements in traditional KPIs, leaders should isolate metrics such as task completion time, accuracy, and economic value generated by digital labor systems. This would make the marginal product of intelligence visible and actionable, preventing AI-driven gains from being lost in aggregate measures.

Moreover, successful implementation requires organizational redesign. Firms must cultivate roles like AI trainers, data curators, and prompt engineers. They must also invest in governance frameworks to ensure AI systems align with business values and performance expectations. Digital labor’s effectiveness hinges on the quality of underlying data and the alignment of human oversight, areas where talent investment becomes paramount.

In this new paradigm, human workers are not rendered obsolete but become pivotal in training, guiding, and scaling AI. Organizations that master this feedback loop, where humans build better AI, and AI enhances human capability, are likely to develop a sustainable competitive advantage.