Krakowski, S., Luger, J., & Raisch, S. (2023). Artificial intelligence and the changing sources of competitive advantage. Strategic Management Journal, 44(6), 1425–1452. https://doi.org/10.1002/smj.3387
Summary
The rapid adoption of Artificial Intelligence (AI) has already begun to disrupt competitive landscapes, with firms either substituting or complementing employees’ cognitive abilities through AI-based tools. In the article, Krakowski et al. apply resource-based view (RBV) principles to investigate how AI changes the foundation of competitive advantage. The study takes a particularly novel approach: it examines chess competitions, where AI-based “engines” have become commonplace in tournaments, radically altering how the game is played and won. By analyzing how competitive outcomes shift across different tournament formats—conventional chess (human vs. human), centaur chess (human plus AI vs. another human plus AI), and engine chess (AI vs. AI, with human input for selection and tuning)—the authors demonstrate two interrelated forces: substitution and complementation of human capabilities by AI.
From Human Capabilities to AI Substitution
In traditional (conventional) chess, human players’ capabilities (like strategic thinking, memory, and pattern recognition) are the primary source of advantage. Extensive research has shown that a high Elo rating (the standardized chess ranking) predicts better performance. Hence, top grandmasters typically outperform lesser-rated opponents.
However, once AI-based “chess engines” entered the scene, this familiar source of advantage began to erode. Engines boast massive computational speed and never tire. This means:
- Substitution of Human Cognitive Skills: The ability of engines to process and evaluate millions of moves per second makes the old skillset of memorizing opening lines and middle-game strategies less critical. The authors found that once AI is available, human ratings (Elo) become far less predictive of who wins – implying that engines can replace a big chunk of what used to be the basis of a superior chess skill.
- Resource Fungibility and Zero Marginal Cost: Another insight is that these AI-based chess engines (e.g., Stockfish, Komodo) are often free or cost very little to replicate. Therefore, in the new AI-enhanced environment, “machine” capabilities become widely available, undermining the advantage a single top player might have had. In other words, it’s much easier (and cheaper) for multiple competitors to access the same high-level engine.
- Convergence of Performance: Since engines rarely make mistakes, chess matches become more accurate. The authors note an increase in drawn games and a decrease in “blunders,” suggesting that the skill disparities among players matter less when they lean on AI to generate top moves.
Complementation: Humans + AI and Human-Governed Engines
Although human players’ traditional skills lost value, a new source of advantage emerged: human–machine capabilities. The authors call this “centaur” capability when humans partner closely with an AI engine during a game, or “human–engine capability” when humans choose, tune, and guide an engine in purely machine-vs.-machine matches. Contrary to what one might expect, humans aren’t simply bystanders. They still exert influence:
- Human–Centaur Capabilities: In centaur chess, humans work with chess engines in real time, consulting the engine’s computations while injecting strategic oversight. This could entail anticipating an opponent’s surprise strategy or harnessing contextual knowledge that a machine alone might not factor in. Interestingly, the authors find that human–centaur capabilities create fresh performance differences. Skills like creativity, intuition about opponents, and the ability to interpret engine suggestions can drive a team of (human + AI) to victory over other (human + AI) pairings.
- Human–Engine Capabilities in Pure Machine Battles: Even if two chess engines face off, humans remain involved in choosing which engine to deploy, setting it up, tuning it, and updating data libraries. These tasks demand different cognitive abilities than the old chess skillset. As a result, a person’s Elo rating might not predict success in an engine-vs.-engine contest. Instead, the best “engine managers” build new competencies—knowing how to exploit vulnerabilities in the AI, diagnosing the code’s weaknesses, or selecting unusual engine configurations that can outsmart a standard approach.
- Evidence of New and Non-Overlapping Skills: The article’s data show that high Elo players might not automatically excel at managing or collaborating with AI. In some cases, being a strong human player can even negatively correlate with the ability to integrate AI-based analysis. Why? Because a top grandmaster’s intuition or fixed mental repertoire can conflict with the AI’s recommendations. Conversely, a mid-level or lesser-rated player who is more flexible and open-minded might exploit the engine’s capabilities better.
Implications for Competitive Advantage
These findings highlight a shift in the sources of competitive advantage once AI enters the scene:
- Destruction of the Old Source: The cognitive skill that once dominated (i.e., raw chess understanding) no longer secures wins in an AI-enabled environment.
- Emergence of a New Source: The new advantage stems from how well humans integrate or orchestrate AI, rather than solely from their original domain-specific expertise.
- Possible Coexistence of Formats: Even though AI has changed the game, “traditional” formats (human vs. human) still exist. But their relative prestige or audience interest may shift over time, mirroring business situations where certain legacy products or services remain but produce diminished profits.
Generalizing Beyond Chess
The authors argue that businesses face similar dynamics. For instance, in medicine, certain AI diagnostics can outperform top physicians at reading scans. That undermines a specialist’s old advantage (long years of medical school and experience), but it also opens a path to new roles—e.g., “medical AI integrators” or “care coordinators”—who might combine personal patient knowledge, interpersonal skills, and creative problem-solving with AI-based data.
In short, the introduction of AI both substitutes and complements humans. Substitution erodes old advantages by making them abundant, while complementation creates novel advantages that hinge on unique human–machine interactions. The key managerial takeaway is that companies should invest in capabilities of orchestration, meaning how human employees interface with and oversee AI-driven systems, rather than relying solely on prior human talents.
Key Lessons
- Substitution Is Real: AI-based machines can quickly match or exceed many “expert” tasks, so old forms of expertise become less distinctive.
- Complementation Opportunity: New forms of human–AI interaction can spur lasting competitive edges—provided they remain rare, hard to copy, and valuable.
- Capability Shift: A star performer in a traditional sense might not be the best at leveraging AI, and some lesser experts become top performers if they excel at adopting, tuning, or governing AI.
- Distributed Technology: Because AI algorithms and software are often open source or easily licensed, they spread fast, meaning no single firm (or individual) monopolizes the tech for long.
- Human Leadership Matters: Selecting, combining, and steering AI systems can create new, durable differences in performance, much like how certain chess players “tune” their engines more creatively.
10 Practical Insights for Business Owners and Managers
- Recognize the Dual Nature of AI
AI can both replace and augment human capabilities. Don’t assume it’s one or the other—plan for both. - Identify “Machine-Orchestration” Roles
Just as “centaur” players emerged in chess, businesses need roles dedicated to tuning, governing, and integrating AI-based tools. - Upskill Beyond the Old Expertise
Being a star in the traditional domain (e.g., medicine, investing, design) won’t guarantee success with AI. Develop new competencies around AI analytics, data interpretation, and creative scenario testing. - Foster Flexibility and Openness
Skilled but rigid experts may ignore or override AI’s suggestions. Encourage a culture where humans check their biases and remain receptive to machine insights. - Leverage AI’s Scale and Accuracy
Simple tasks like data analysis or shortlisting can be taken over by AI to free up human employees for broader tasks like strategy, communication, or creative ideation. - Concentrate on Complementary Strengths
Humans shine at empathy, holistic context, or persuading stakeholders, while AI excels at large-scale pattern detection. Combine these in ways competitors find hard to replicate. - Invest in Rare Collaboration Skills
Because AI is widely available, the real edge is in how your teams collaborate with it. This skill is scarcer and can’t be instantly cloned, leading to a more defensible advantage. - Be Prepared for Value Shift
If your organization used to excel on deep domain expertise, be aware that domain skill alone might decline in value. Shift resources toward AI-human orchestration and advanced problem-solving. - Rethink Recruiting and Team Composition
Look for people with both domain knowledge and the adaptability to harness AI. You might need team “hybrids” who bridge tech-savviness and subject-matter insight. - Experiment to Find the Right Format
In chess, conventional, centaur, and engine events coexist. In business, some tasks remain purely human, others purely automated, and others involve close AI-human collaboration. Try pilot projects and let data guide you on which approach works best for each task.
Closing Thought
Krakowski et al. show that adopting AI transforms competition by simultaneously rendering old human-centric advantages obsolete and creating fresh opportunities for human–AI synergy. In a business context, leaders should anticipate a reshaping of competitive advantage as AI substitutes certain established capabilities while opening up new domains of complementarity and orchestration. Those who adapt quickly—by cultivating teams and processes that smartly blend AI’s computational power with uniquely human creativity and insight—will thrive in the AI-enhanced marketplace.