Redefining Business Strategy in the Age of AI

From Scarcity to Abundance of Expertise

With the fast-changing business world, artificial intelligence is transforming the economics of expertise at its very roots. Expertise, once a limited resource that organizations painstakingly developed and guarded, is now becoming ever more widespread and available. This change necessitates a complete revision of business strategy, organizational design, and competitive advantage.

As Harvard Business School Professor Karim R. Lakhani and colleagues in their recent Harvard Business Review article write, “AI is changing the cost and availability of expertise, and that will fundamentally alter how businesses organize and compete” (Lakhani et al., 2025). The change is not only a marginal change but a paradigm shift in how companies develop and capture value.

The Evolution of the Firm: From Coase to the AI Era

To get a sense of the scale of this change, we have to go back to the underlying theory that has governed our knowledge about why companies are the way they are. In 1937, the British economist Ronald Coase wrote his now legendary paper “The Nature of the Firm,” asking a disarmingly straightforward question: If markets are so good at managing resources, why do companies even exist?

His response fundamentally transformed economic thought; firms exist due to the existence of costs associated with utilizing the market mechanism what he referred to as “transaction costs” (Coase, 1937).

These transaction costs are incurred in searching for suppliers, negotiating contracts, coordinating effort, safeguarding intellectual property, and monitoring performance. When these costs outweigh the benefits derived from market transactions, firms decide to internalize activities. This seminal observation won Ronald Coase the Nobel Prize in 1991 and has influenced our understanding of organizational boundaries for nearly a century.

In the industrial era, Coase’s reasoning accurately predicted the growth of firms. Companies grew larger, building complex hierarchical organizations to manage a growing number of internal capabilities. General Motors in Alfred Sloan’s day was a textbook case, integrating backward from finished cars to raw materials. IBM also built massive internal capabilities in hardware, software, and services.

But as MIT economist Tom Malone articulated in his 1987 article “Electronic Markets and Electronic Hierarchies,” information technology started to alter this calculus by lowering the costs of coordination (Malone et al., 1987). The internet’s advent further accelerated the process, making it possible to build sophisticated global supply chains and new forms of collaboration. Vertically integrated firms gradually turned into “virtual enterprises,” depending increasingly on partners for tasks that were once handled internally.

AI represents the next underlying inflection point on this evolutionary path. Unlike previous technologies, which were primarily concerned with reducing the cost of coordination, AI is democratizing knowledge itself. The new capability of modern AI systems to encode, transmit, and apply expertise at scale changes the economics of expertise that have underpinned organizational form since Coase’s time.

The Triple Product of AI Integration

There are three significant ways AI is changing business: reducing cost and time, enhancing workforce productivity, and reallocating strategic resources  their “triple product” (Lakhani et al., 2025).

First, artificial intelligence allows significant operational effectiveness through process change. In the past, organizations looking to reduce costs often relied on outsourcing or offshoring; however, these practices usually required moving whole processes. Artificial intelligence changes this calculus as it enables organizations to utilize outside expertise on a specific task within a process. JPMorgan Chase, for instance, implemented an AI system named COIN (Contract Intelligence) which reviews documents for loan agreements in seconds, a process that took lawyers and loan officers 360,000 hours a year (Son, 2017). This selective automation maintains the integrity of the overall process while saving costs on individual tasks significantly.

Second, artificial intelligence enhances the skill level of the workforce at various levels of performance. “As firms adopt AI assistants, the assistants will in effect put at least a minimum level of expertise in the hands of every worker who uses them, enabling that person to do a better job” (Lakhani et al., 2025). Consider how GitHub Copilot transforms coding productivity by suggesting entire functions as developers write code, or how AI-powered medical diagnostic software helps doctors identify conditions that they might otherwise miss. Such machines don’t replace human talent but augment it, enabling workers at all skill levels to perform at greater levels.

Third, and most strategically important, AI enables companies to reconsider fundamentally how they deploy resources. Companies can identify their actual differentiators areas where they possess world-class abilities that create unique value and double down on those and utilize AI for non-core functions.

Reconfiguration of organizational boundaries is consistent with the findings of John Hagel and John Seely Brown in their research on institutional innovation. They have argued that “Institutions that can drive accelerated learning will be the most likely to thrive in today’s environment of exponential technology change and market uncertainty” (Hagel & Brown, 2013).

Job Polarization and the Evolving Nature of Work

The influence of artificial intelligence on organizational boundaries and expertise is witnessed through its influence on labor markets. As documented by MIT economist David Autor, information technology has been a central driver of the polarization of job opportunities (Autor, 2015). Workers in abstract task-intensive occupations in higher education are exposed to substantial complementarities with technology, whereas those in routine task-intensive occupations are displaced.

Autor’s study reveals that between 1979 and 2012, job growth was focused in both high-skill, high-wage jobs and low-skill, low-wage jobs, while middle-skill jobs proportionally declined (Autor, 2015). This pattern reflects the differential impact of technology on various kinds of tasks. Jobs involving routine, codifiable tasks cognitive, like bookkeeping, or manual, like assembly line work are especially susceptible to automation. Conversely, jobs that require problem-solving, flexibility, and interpersonal skills have proven more resilient.

What Autor has referred to as “Polanyi’s paradox”  referencing philosopher Michael Polanyi’s observation that “we know more than we can tell”  is responsible for this trend (Polanyi, 1966). Most human capability relies on tacit knowledge that has been difficult to codify and automate. This distinction between explicit and tacit knowledge has historically protected certain professions from technological replacement.

But artificial intelligence is starting to overcome this boundary with machine learning developments. Instead of needing the explicit programming of rules, these systems learn patterns in data and so may sidestep some of Polanyi’s paradox. Political Economy of Communication implications are deep: businesses must rethink what elements of expertise are still distinctly human and what elements are being commoditized by artificial intelligence.

Autor predicts that, instead of going on ad infinitum, labor polarization will morph with the emergence of new hybrid occupations. He envisions the development of what Harvard economist Lawrence Katz has referred to as “new artisans”  individuals that blend technical and interpersonal skills not readily replicable by artificial intelligence (Autor, 2015). This perspective indicates to businesses a range of strategic choices in restructuring work to exploit the complementarities between human and artificial intelligence.

Strategic Decision Model for the AI Era

The new economics of expertise requires a fresh strategic agenda. Companies must systematically review their portfolio of expertise, distinguishing between differentiating capabilities that create unique value and those that, while necessary, do not drive competitive advantage.

This analysis starts with three pertinent questions that Lakhani and co-authors suggest:

To begin with, what are the issues that your company solves today for customers that they will ultimately solve for themselves using AI? Consider how TurboTax changed tax preparation or how graphic design software such as Canva democratized design. AI will accelerate this trend across sectors, compelling companies to move up the value chain or be disintermediated.

Second, what expertise needs to develop most quickly to remain ahead of AI capabilities? Goldman Sachs is a good example. When computer algorithms commoditized old-style trading expertise, the company moved into more sophisticated financial engineering and advisory work. Companies need to constantly reassess what expertise continues to be differentiating as AI capabilities improve.

Third, what are the assets that can be created or improved to enhance competitiveness as artificial intelligence evolves? Data assets, confidential algorithms, and special customer relationships may prove more sustainable sources of competitive advantage than plain expertise by itself.

These questions form the basis of a strategic sorting activity that entails classifying expertise in order to make decisions about allocating resources. The resultant framework could include:

1. Core differentiators: consist of expertise that creates unique value and resists commoditization.
2. Evolution candidates: Knowledge that remains applicable but needs to evolve as AI changes
3. AI augmentation targets: Areas where human abilities can be significantly enhanced by AI
4. Outsourcing applicants: Tasks in which in-house skills no longer deliver competitive edge

Microsoft’s reinvention under Satya Nadella’s direction is a textbook case of this approach. Recognizing that traditional software capabilities were increasingly commoditizing, Microsoft made a strategic shift to cloud infrastructure and AI services while simultaneously building out complementary capabilities in developer ecosystems and enterprise relationships. This strategic shift has generated enormous value, with Microsoft’s market capitalization growing from around $300 billion in 2014 to more than $3 trillion as of 2025.

Case Studies: Winners and Losers in the AI Transition

The shift to AI-rich expertise is producing obvious winners and losers in every industry. Look at the diverging fortunes of two financial services companies: Capital One and a typical regional bank (which we’ll keep anonymous for this conversation).

Capital One saw early that artificial intelligence was going to disrupt lending decisions. Instead of viewing credit analysis as a core competency to defend, they repositioned their brand as a tech company that happens to lend. They made big investments in artificial intelligence capabilities, developed proprietary data assets, and redesigned workflows to take better advantage of these technologies. The payoff has been improved loan performance, lower operating costs, and the ability to profitably serve previously unprofitable customer segments.

In contrast, the regional bank viewed artificial intelligence mainly as a cost reduction tool and adopted discrete solutions without reconsidering the core of their expertise model. They persisted in investing in conventional credit analysis capabilities while lagging in data infrastructure and analytical savoir-faire. As lending decisions became more automatic and commoditized, their unique expertise dwindled, resulting in margin compression and loss of market share.

We are seeing similar patterns in the healthcare industry. Institutions like the Mayo Clinic are tactically incorporating artificial intelligence into clinical workflows while simultaneously highlighting their unique expertise in complicated cases and patient relationships. They are leveraging AI to handle routine diagnosis and treatment planning while focusing human expertise on high-complexity cases and compassionate care that is still outside the realm of artificial intelligence.

These examples show a typical pattern: successful organizations don’t just apply AI within traditional expertise models; rather, they fundamentally rethink what kinds of expertise generate new value in an AI-saturated environment.

Conclusion: Preparing for the Future of Business

The move from limited to unlimited expertise is a fundamental change in the business basics. Just as Coase’s theory of the firm clarified organizational boundaries during the industrial age, new frameworks are necessary for the AI age. The lines between firms and markets, humans and machines, core and peripheral expertise are all being redrawn.

Successfully managing this shift needs more than step-by-step modification. It needs revolutionary reconsideration of how companies create and capture value. Those likely to win are the ones able to analytically examine their portfolio of competencies, strategically invest resources to propel emerging differentiators, and develop complementary assets that extend competitive advantage.

As MIT economist David Autor has observed, the interplay between human labor and technology is marked by a subtlety that defies simple substitution (Autor, 2015). Automation in the past has worked to augment human labor, even as it displaces specific tasks. Thus, the business challenge for leaders is to discover new modes of complementarity between human and artificial intelligence—capturing the value of each in its own sphere while, at the same time, creating new sources of value at their intersection. Organizations of the future will have more permeable boundaries, greater focus on their investments in knowledge, and more capability to tap external resources than organizations today. Though the transformation will be anything but easy, those that handle it successfully will emerge with enhanced strength, resilience, and improved capability to add value in an era where expertise is in abundance.

References

Autor, D. H. (2015). Why Are There Still So Many Jobs? The History and Future of Workplace Automation. Journal of Economic Perspectives, 29(3), 3-30.

Coase, R. H. (1937). The Nature of the Firm. Economica, 4(16), 386-405.

Hagel, J., & Brown, J. S. (2013). Institutional Innovation. Deloitte University Press.

Lakhani, K. R., Yerramilli-Rao, B., Corwin, J., & Li, Y. (2025). Strategy in an Era of Abundant Expertise. Harvard Business Review.

Malone, T. W., Yates, J., & Benjamin, R. I. (1987). Electronic Markets and Electronic Hierarchies. Communications of the ACM, 30(6), 484-497.

Polanyi, M. (1966). The Tacit Dimension. University of Chicago Press.

Son, H. (2017, February 28). JPMorgan Software Does in Seconds What Took Lawyers 360,000 Hours. Bloomberg.