Organizations Flourish on the Edge of Chaos: The Timeless Strategy of Go

In a game of Go, two players compete for territory by placing black and white stones on a 19×19 gridded board. By Woohae Cho. (NYT)

When I developed my thesis on the organizational genome, I drew on principles from the natural sciences. For this article, I examine the principles of the game of Go and the implications of AI AlphaGo, with a particular focus on how AI has influenced strategic thinking.

Context

Like master Go players surveying an intricate board, scientists from various fields, including biology, chemistry, physics, and beyond, have studied the patterns of systems theory over the past decade. The Santa Fe Institute exemplifies this path. The natural sciences' findings, when linked with organizational design, reveal a strategic landscape where twenty-first-century organizations can secure a foothold and thrive, much like building “Thickness” in Go, an unassailable foundation that enables future moves. Thickness from an organizational viewpoint refers to robust foundations (strong culture, capabilities, relationships). This perspective challenges managers to shift from viewing organizations as predictable (ergodic) machines to seeing them as dynamic systems, akin to a Go game played across deep time, where stones (agents) shape the board (environment) through interaction and strategy.

What is Go?

Go is an ancient board game that originated in China over 2,500 years ago. It is played on a grid, typically 19×19, where two players alternate placing black and white stones to capture territory and surround their opponent’s stones. Despite its simple rules—players take turns placing stones on intersections to control more area than their opponent—Go is renowned for its profound strategic depth. Often compared to chess. Go offers exponentially more possibilities due to its larger board and fewer restrictions on moves, requiring players to balance short-term tactics with long-term strategy. In Go, victory comes not from capturing pieces as in Chess, but from controlling territory through robust formations and latent opportunities (Aji). 

Similarly, organizations must evolve beyond short-term, tactical responses—akin to Chess’s direct attacks—to strategic positioning that ensures adaptability. This article leverages the Go metaphor to explore complex adaptive systems (CAS), enabling managers to build resilient and evolvable organizations capable of navigating the complexities of the modern world.

The AlphaGo Revolution: When AI Redefined Strategic Thinking

In March 2016, DeepMind’s AlphaGo achieved what many thought impossible: defeating Lee Sedol, one of the world’s greatest Go masters, in a five-game match. This victory wasn’t just a technological milestone—it fundamentally challenged how we understand strategy itself. AlphaGo’s approach revealed unconventional moves that initially seemed illogical to human masters but proved strategically sound when viewed across the entire game arc.

Most remarkably, AlphaGo didn’t simply calculate faster than humans; it developed novel strategic patterns that expanded the very conception of optimal play. Move 37 in Game 2—a shoulder hit on the fifth line that professional commentators initially dismissed as a mistake—demonstrated how AI could transcend traditional strategic frameworks to discover new forms of thickness and aji.

Implications for the Evolution of Organizational Design

The AlphaGo breakthrough offers profound insights for organizational design:

Pattern Recognition Beyond Human Intuition: Just as AlphaGo identified strategic patterns invisible to human masters, AI-augmented organizations can recognize market opportunities, operational inefficiencies, and competitive threats that escape traditional analysis. Organizations must develop systems that can process vast amounts of data to identify emergent patterns—their version of Move 37.

Hybrid Intelligence Architecture: Lee Sedol’s victory in Game 4 came from a brilliant human move that exploited AlphaGo’s blind spot, demonstrating that human creativity remains irreplaceable. The future of organizational design lies not in replacing human judgment with AI, but in creating hybrid systems where artificial intelligence enhances human strategic thinking. Like Go players who now train with AI to expand their strategic repertoire, organizations need structures that amplify both human insight and computational power.

Dynamic Strategy Evolution: AlphaGo’s self-play training—where it played millions of games against itself—suggests that organizations should create internal environments for continuous strategic experimentation. Adaptive organizations require processes for ongoing strategic iteration and learning from simulated scenarios, not inflexible cycles of strategic planning.

Redefining Competitive Advantage: Before AlphaGo, the Go strategy was constrained by centuries of human wisdom. Post-AlphaGo, the strategic landscape exploded with new possibilities. Similarly, organizations must prepare for competitors who leverage AI to discover entirely new business models, operational approaches, or customer engagement strategies that seem counterintuitive but prove superior.

The Emergence of Algorithmic Thickness: AlphaGo built thickness not through traditional territorial control, but through deep positional understanding that created multiple future possibilities. Modern organizations need to develop “algorithmic thickness”—robust data architectures, AI capabilities, and learning systems that create strategic options even when specific future scenarios remain uncertain.

The AlphaGo victory reminds us that in both Go and organizational strategy, the most profound advances come not from playing the existing game better, but from reimagining what the game itself could become.

Analogy Between Experiments in “Playing Go with Darwin” and the Game of Go

In the article Playing Go with Darwin, evolution is likened to a strategic game of Go, where organisms “play” by adapting to their environment through genetic changes. The experiments discussed, such as those involving protein evolution in E. coli, show how natural selection can enhance an organism’s “evolvability”—its capacity to adapt to future challenges. This process mirrors key strategies in Go. For example, in the game, players build “Thickness”—robust, unassailable positions on the board—to create “Aji,” or latent potential for future moves. Likewise, the experiments demonstrate that strong selection in proteins can flatten the “fitness landscape,” thereby reducing constraints and allowing for greater variability, which facilitates faster adaptation. Just as a Go player might sacrifice a few stones to secure a stronger position later, evolution may forgo immediate benefits to build resilience or adaptability over time. This analogy highlights how both evolution and Go involve a delicate balance of immediate tactical decisions and long-term strategic foresight, providing a powerful lens for understanding organizational dynamics.

Working Definitions

In the “organizational game” of Go, key terms take on new meaning:

Agent: Agents are the “stones”—employees, managers, or stakeholders, each with unique strategies and positions. Their interactions shape the board, creating emergent patterns.

System: A system is the board itself—a network of agents whose collective moves influence the whole.

Adaptation: Adaptation mirrors a Go player adjusting to the shifting board state, building Thickness (robustness) to seize future opportunities (Aji).

Self-Organization: Like Go players crafting patterns without a central referee, agents self-organize into effective structures through local interactions.

Emergent Property: These are the “territories” claimed—outcomes like innovation or culture that emerge from agent interplay.

Complexity: A complex system is a mid-game Go board, poised between order and chaos, fostering adaptability.

Selection: Selection is the survival of strong positions—effective traits or practices persist under pressure.

Edge of Chaos: Organizations thrive at this Go-like balance—structured yet flexible.

Co-evolution: As in Go, where each player’s move shapes the other’s strategy, organizations co-evolve with their environment.

Thinking Differently

Historically, we've worked with organizations like Chess, which have clear hierarchies and immediate objectives. However, today's economy resembles Go, where indirect influence and foresight prevail over linear tactics. In Go, players build Thickness to secure territory, creating Aji for future plays. Organizations must foster robustness by implementing diverse teams or flexible structures, not only to ensure their survival but also to unlock the potential for evolvability.

The AlphaGo revolution amplifies this shift. Just as Lee Sedol had to adapt his centuries-old strategic framework when faced with AlphaGo's unconventional moves, modern organizations must prepare for competitors who leverage AI to discover entirely new strategic approaches. The question is no longer just about building human thickness, but about developing hybrid intelligence systems that can both recognize patterns beyond human perception and maintain the creative flexibility that allowed Lee Sedol to find AlphaGo's weakness in Game 4.

The Convergence of Systems Thinking with Complexity Theory

In Go, simple rules yield immense complexity. Organizations, as CAS, follow suit: agents operate with rules, their interactions spawning emergent behaviors. Like Go players using Gote (defensive moves) to build Thickness, organizations must foster robustness to pursue Sente (offensive opportunities). The Playing Go with Darwin analogy reinforces this: strong selection enhances evolvability, while entrenched practices solidify complexity.

This emergence of unpredictable complexity from simple rules reflects a fundamental principle that extends beyond organizational theory. A new theory suggests that the fundamental laws of physics are not “complete” in the sense of providing all the information necessary to comprehend natural phenomena; instead, evolution—biological or otherwise—introduces functions and novelties that could not be predicted in principle from physics alone. This perspective sheds light on why organizations, like living systems, exhibit emergent properties that transcend their individual components. Just as biological evolution creates capabilities that cannot be deduced from chemistry and physics, organizational evolution generates strategic advantages, cultural innovations, and adaptive responses that emerge from—but cannot be predicted by—the formal rules and structures that govern individual agents.

The AlphaGo era adds another dimension: the convergence must now account for algorithmic agents alongside human ones. Organizations exist in an ecosystem where some competitors operate with AI-enhanced pattern recognition and strategy generation. This creates a new form of co-evolution, where human creativity and artificial intelligence must work in synergy to build a sustainable competitive advantage. The emergence of AI strategies, such as AlphaGo's Move 37, exemplifies how novel functions can arise that were not predictable from the original programming rules—a perfect illustration of how evolution, whether biological, organizational, or algorithmic, transcends its foundational constraints to create genuine novelty.

Functional Examples

• Systems Thinking: In Go, players see the whole board. A car manufacturer must integrate skills systemically, like connected stones.

• Adaptation: Active adaptation builds Aji, seizing the initiative strategically.

• Functional Design vs. CAS: A CAS view prioritizes long-term positioning over quick wins.

Process for Changing Mechanical Mental Models in Human Capital

• Establish Urgency: Highlight the need for a Go mindset.

• Develop Diverse Agents: Enrich the organizational board.

• Utilize Feedback Loops: Strike a Balance Between Stability and Innovation.

• Build Commitment: Strengthen Thickness across boundaries.

• Continuously Improve: Build Aji for future plays.

Conclusion

In Go, triumph lies in controlling territory through strategic placement, not just capturing stones. The AlphaGo revolution demonstrated that even ancient strategic wisdom can be revolutionized through new forms of intelligence. Organizations must adopt this expanded mindset, building robustness and evolvability that encompasses both human creativity and artificial intelligence, to dominate their competitive landscape.

By viewing themselves as CAS—dynamic, adaptive, and strategically poised in an AI-augmented world—managers can transcend mechanical models, securing a future where adaptability and hybrid intelligence become the ultimate victory. The lesson from AlphaGo is clear: the most profound advances come not from playing the existing game better, but from reimagining what the game itself could become.

References

Note: Adapted from my original thesis, taking the main arguments of the author and adapting them to the Go principles.

Bennis, Warren. Beyond Bureaucracy. 1993. Explores moving past Chess-like bureaucracy to Go-like adaptability.

Durlauf, Steven N. “Economic Complexity.” 1997. Highlights nonlinear interdependencies, akin to Go’s complex board.

Ackoff, Russell L. The Democratic Corporation. 1994. Advocates systemic design, mirroring Go’s strategic depth.

Gharajedaghi, Jamshid. System Thinking. 1999. Frames complexity as a Go-like rhythm of order and adaptability.

Drucker, Peter. Management Challenges. 1999. Calls for constant change, akin to Go’s evolving board.

Forrester, Jay W. “Counterintuitive Behavior.” 1995. Notes social systems’ complexity, like Go’s unpredictable patterns.

Senge, Peter. The Fifth Discipline. 1994. Emphasizes learning as Aji for organizational evolution.

Tapscott, Don. The Digital Economy. 1996. Describes networked intelligence as a Go-like CAS.

Krakauer, David. “Playing Go with Darwin.” Nautilus, December 15, 2020. This article explores the parallels between evolutionary strategies and the game of Go, using experiments on protein evolution to illustrate how natural selection fosters long-term adaptability, much like building strategic depth on the Go board.

Silver, David, et al. "Mastering the game of Go with deep neural networks and tree search." Nature 529.7587 (2016): 484-489. The seminal paper describing AlphaGo's architecture and its implications for strategic thinking beyond traditional human frameworks.

Final Remarks

I utilized AI platforms, including Gemini, Claude, ChatGPT, and Grammarly, to expedite research while ensuring clarity and logical flow. My aim in using these tools was to verify information across multiple sources and validate it through academic databases and collaborations with equity firm analysts.