Systems Thinking Series 1: When everyone is right, but everything still goes wrong!

Take a moment to think back to how we were taught to solve problems in school. We were told to break them down. Isolate the variables. Solve for ‘x’. Find the cleanest path to the right answer. We were told, “If you get the parts right, the whole will follow.” 

From an early age, our education system conditions us to divide problems into discrete parts — physics, chemistry, history, math. Even within subjects, we’re trained to solve isolated equations, analyse standalone texts, memorize historical events without delving into causality. That’s not inherently wrong. In fact, reductionism is powerful — it gave us π and calculus, Newtonian mechanics, antibiotics, and modern programming. 

But here’s the catch: we mastered differentiation but forgot integration. We break things down but rarely reassemble the parts to examine the system holistically. Reductionism is neat. It’s tidy. It works, until it doesn’t. Because the real world has complex systems, and systems behave differently.

Our school education doesn’t train students to ask:

• How does this connect?

• What changes when two pieces interact?

• Who or what else is affected by this?

• What might happen later because of doing this now?

As a result, we produce brilliant specialists who can optimize a subsystem but often miss the bigger picture. A software engineer might optimize a database query but overlook how it affects product experience. A policymaker might solve for traffic but worsen pollution. A procurement team might reduce unit costs and inadvertently increase total cost of ownership through poor quality or excess complexity. It’s no surprise, then, that great technocrats don’t always make great leaders. Solving for a part isn’t the same as leading a system.

Systems have a mind of their own

In The Right Kind of Wrong, Amy Edmondson writes “Systems exhibit synergy. The whole is more than the sum of the parts, and the behaviour of the whole can’t be predicted by the behaviour of the parts examined separately. Only by considering the relationships between parts can you explain a system’s behaviour.” These words are not just insightful; they are painfully relevant in today’s world.

Think about it:

• Climate change: We can’t solve it by just looking at emissions from cars, or factories, or agriculture in isolation.

• Healthcare: Treating symptoms won’t fix the root causes — poor nutrition, chronic stress, mental health, social inequity.

• Organizational culture: We can’t build collaboration by optimizing each team independently. We must redesign the relationships between them.

  
  

In all these cases, the magic (and the mess) lives in the interconnections.

Systems Thinking in the age of AI and complexity

Today’s world is more connected, more automated and more complex than ever before. Which makes systems thinking not optional but existential. Think about it: the human brain doesn’t solve problems by isolating ‘x’. It’s a dense network of billions of neurons where even the simplest task requires hundreds of thousands of connections working together. It’s a system in motion, not a set of isolated switches.

The biggest challenges of our time — geopolitical volatility, supply chain fragility, pandemics, major accidents, wealth disparity, terrorism — are all systemic problems. They can’t be solved with isolated interventions. They require: 

• Nonlinear thinking

• Multi-stakeholder modelling

• Long-term, dynamic planning

• Feedback awareness

• Tolerance for ambiguity 

And yet our systems continue to reward solving tidy, bounded problems with clean answers, rather than mapping dynamic systems in motion.

So, what should we do? 

1. Teach interconnected thinking early

Systems maps, causal loops, and simulation-based learning shouldn’t be electives — they should be core curriculum. Because real-world problems don’t show up neatly labelled by subject. By exposing students to the complexity of interconnected systems early on, we build cognitive habits that embrace relationships, not just results. It teaches them to trace consequences, anticipate ripple effects and think in terms of flows, not just facts.

2. Redefine assessment

Move beyond right vs. wrong. Reward quality of reasoning, diversity of perspective, and depth of insight across disciplines. Too often, we assess people on whether they followed the ‘expected’ path — rather than whether they explored the right problem in a meaningful way. In a systems world, the most valuable skill isn't arriving at a perfect answer, it’s navigating ambiguity with clarity and being able to explain your thinking. We need to recognize not just what someone concluded, but how they got there, and who or what they considered along the way.

3. Simulate real complexity

Case studies, real-world modelling, even systems games like The Beer Game should replace worksheets and drills. These activities build intuition that no formula can teach. They reveal how delays, dependencies, and decision loops play out in real life. They show students that even when everyone acts rationally, systems can behave irrationally. By confronting learners with messy, multi-variable challenges that evolve over time, we train them to think like designers, strategists, and systems stewards, not just answer machines. 

4. Ask better questions 

Systems thinking doesn’t always give clean answers. It gives better questions. Elon Musk’s famous words, “Very often the issue is understanding what question to ask. And if you can properly frame the question, then the answer is the easy path”, sounds prophetic. Because Prompt Engineering, which is at the heart of AI, is the art of asking the right questions and designing the right prompts in a way that AI is able to give clean and accurate answers. 

Designing the future 

The industrial-age education system trained people to operate machines and manage hierarchies. But the 21st century demands something very different: People who can see complexity, ask the right questions, get comfortable with uncertainty and design for emergence. Systems don’t wait for us to figure them out. They react. They evolve. To design a better future, we must first learn to see — not just the parts, but the whole.