The AI Divide: A Tale of Two Capabilities

The national conversation around artificial intelligence (AI) has focused on job displacement, skill requirements, and technological advancements. However, these discussions only scratch the surface of a more fundamental issue that has far-reaching implications for organizations, governments, and individuals.

At its core, the AI revolution is not just about developing smarter models or leveraging new technologies; it’s about reorganizing work around them. This shift from invention to implementation requires specific skills and capabilities that are in short supply, particularly among smaller institutions and organizations with limited resources.

Consider a regional health system struggling to process prior authorization requests with the aid of AI models. Human judgment and oversight are still necessary to ensure accuracy and accountability. But who is responsible when errors occur? Who decides which submissions need manual review, and how do we measure the impact of these new technologies?

Policymakers and industry leaders have focused on secondary issues – like which jobs AI will change, which skills will matter, and which industries will move fastest. However, these questions are less pressing than concerns about organizational capability.

The market is already speaking to this issue. Job postings for companies building frontier AI models often include roles like forward-deployed engineer, AI transformation lead, and solutions architect. These positions don’t focus on making models smarter; they concentrate on helping customers use existing technology effectively.

As AI becomes a commodity, institutional capability will be the deciding factor in competitive advantage and institutional inequality. The next AI divide won’t separate organizations by who holds the most powerful model but by those that can put models to work and keep them working.

The people who sit between technology and operations – translators, workflow designers, change managers, evaluators – are the emerging AI deployment workforce. Without these individuals, AI will become another resource used strategically by well-capitalized organizations while everyone else struggles to make it work.

Economists have documented that general-purpose technologies deliver little measured productivity at first because their payoff depends on complementary intangible investments. Realizing such a technology’s potential requires a fundamental rethinking of the organization of production itself – a process that involves significant changes in managerial experience, worker retraining, and process innovation.

One obvious objection is that these problems will be temporary as models become easier to use and agents manage their own integration. However, governance questions like accountability when systems err and what evidence counts as proof of impact do not get easier; they require a more profound understanding of organizational capability.

As we move forward in this AI revolution, it’s essential to recognize that the real challenge is not about developing new technologies but about reorganizing work around them. We must acknowledge that institutional capability is the key differentiator between successful and unsuccessful organizations – and that this gap will only widen without a concerted effort to address it.

The AI divide is not just about technological advancements; it’s about the people, processes, and governance structures in place to support these innovations. By focusing on organizational capability rather than model sophistication, we can create a more inclusive and equitable future for all stakeholders – one where the benefits of AI are distributed fairly among organizations, governments, and individuals alike.