Executive summary
In the latest episode of our in other words podcast, political scientist Dr. Brian Klaas argues that control is a comforting myth, and that leaders make better decisions when they accept randomness as the baseline and build for resilience, not certainty. In this episode, Klaas explains how small causes can produce outsized effects, why our brains over-detect patterns, and how executives can shift from a control mindset to one of influence, experimentation, and trust. Phrase CEO Georg Ell joins to translate these ideas into leadership practice for complex, global organizations.
Brian illustrates the stakes with stories that show how tiny variables can redirect history and forecasting, from a 1926 Kyoto visit that later influenced a World War II targeting decision, to Edward Lorenz’s discovery that rounding error in early weather models could flip clear skies into storms. The lesson is not to give up on strategy, but to stop pretending that perfect prediction is possible in open, interlinked systems. Leaders should design organizations that can bend, not break.
The conversation moves from theory to application: how to audit decision processes rather than outcomes, why trust is the real buffer during shocks, and how to operationalize agility by balancing planned campaigns with room for reactive moves. Brian frames resilience as a choice, not a buzzword, and argues that over-optimized systems become brittle, citing supply chain fragility as a cautionary tale. He urges leaders to create space for safe-to-fail experiments and to normalize change so teams do not panic when the unexpected arrives.
On AI, Brian distinguishes low-risk closed uses from higher-risk open systems that interact with people and other AIs. The takeaway for the C-suite: AI will be transformative, but it will not repeal chaos; set realistic expectations, build guardrails, and invest in trust, resilience, and experimentation.
Embracing chaos: why leaders must rethink control, resilience, and trust
Executives often rise to the top on the strength of their ability to plan, predict, and control. Spreadsheets, strategies, and forecasts promise a future that bends to vision and discipline. But what if this is the most dangerous myth in leadership?
Political scientist Dr. Brian Klaas, professor at University College London and author of Fluke: Chance, Chaos, and Why Everything We Do Matters, argues that randomness isn’t the exception in business and politics: it’s the rule. Leaders who cling too tightly to control, he says, end up brittle and blindsided. Those who build for resilience, agility, and trust, by contrast, position their organizations to thrive in a turbulent world.
In the latest edition of the In other words podcast, Brian lead a wide-ranging conversation with Phrase CEO Georg Ell and host Jason Hemingway, pulling apart the illusion of control, and illustrating its dangers with stories from history and science, and offering a blueprint for leadership that doesn’t just survive chaos but actively leverages it.
“The myth of control is very seductive. But randomness is a constant. Leaders who pretend otherwise are setting themselves and their organizations up for failure.” – Brian Klaas
The illusion of control
Brian has spent his career studying political systems and the role of chance in shaping events. What he has found is unsettling: behind almost every big historical turning point lurks a small, random event;a hinge moment whose significance only becomes clear in hindsight.
One story he tells comes from Zambia. A coup attempt in the 1990s hinged on a single moment: a soldier trying to grab the trouser leg of the army commander as he scrambled over a wall. The soldier’s grip slipped. The commander escaped, raised the alarm, and the coup collapsed. A second earlier, a tighter grasp, and the government would likely have fallen. “That’s the scale of detail chaos theory makes us confront,” Brian explains. “Tiny actions, overlooked at the time, can redirect the course of nations.”
He pairs that with more famous examples. In Kyoto in 1926, an American official vacationed in the city, fell in love with its culture, and two decades later — as U.S. Secretary of War — argued successfully to spare it from atomic bombing. In 1961, MIT meteorologist Edward Lorenz discovered chaos theory when a rounding error in a weather model produced wildly divergent forecasts: the butterfly effect.
“Business leaders love control,” Brian says. “But these stories remind us that even the most sophisticated systems are at the mercy of randomness. A single tiny variation — a delayed shipment, a regulatory change, a miscommunication — can set off consequences you can’t foresee.”
Jason Hemingway pressed him on the corporate parallel. “So are five-year strategic plans just works of fiction?”
Brian doesn’t hesitate. “When you draw up a rigid plan, you’re assuming the world will stay still. It won’t. The only certainty is that something will come along to derail it. The leaders who succeed aren’t those who guess the future correctly. They’re the ones who build organizations that bend rather than break.”
Georg added the executive perspective: “Strategy matters. But if you treat the plan like a straitjacket, you’ll suffocate the organization. Reality will throw curveballs. Leaders who can’t pivot quickly find that the plan itself becomes the liability.”
“Control nothing, influence everything”
Brian distills the mindset shift into one striking phrase: control nothing, influence everything. Leaders can’t dictate outcomes in chaotic systems, but every small choice — a follow-up email, a hiring decision, a cultural signal — ripples out in unpredictable ways.
“It’s a liberating idea,” Brian argues. “You don’t have to pretend you control the uncontrollable. But you do need to recognize that every decision carries influence, often in ways you’ll never fully see.”
For Georg, that message is empowering: “We talk a lot at Phrase about leadership being a behavior, not a title. If you act with intent — smile at someone in the corridor, give tough feedback, share context — those ripples can uplift the whole company. It’s empowering to realize small choices matter.”
Systems, not superheroes
Brian is wary of another myth: that history is driven by brilliant individuals. His research suggests otherwise.
“Overconfidence in individual leaders is dangerous,” he says. “When everything depends on one figure, the system is brittle. If that person falters, everything collapses. The societies and organizations that survive shocks are the ones with resilient systems.”
He points to political examples where authoritarian regimes crumbled overnight when one leader fell, compared with democracies where institutions absorbed shocks and carried on. “It’s not about the hero at the top,” Brian argues. “It’s about whether the system has distributed strength.”
The lesson for business leaders is clear. Superstar CEOs make headlines, but organizations built on their charisma alone are fragile. A resilient company is one where authority is distributed, teams are empowered, and systems are designed to continue functioning even when leaders stumble.
Georg echoed this from his own leadership experience. “Distributed teams don’t need a superhero CEO. They need context and empowerment. My job isn’t to make every decision. It’s to set vision and create structures so the right decisions can be made at every level.”
Why leaders misread patterns
If control is an illusion, pattern-detection can be just as dangerous. Brian warns that humans are wired to over-detect patterns, a trait inherited from our evolutionary past. In prehistoric times, mistaking wind in the grass for a predator wasted calories, but mistaking a predator for wind got you killed. Today, that bias leads us to infer meaning from noise.
“In business, it’s tempting to believe one customer complaint signals a systemic issue, or that one viral hit proves your strategy flawless,” Brian says. “But often, we’re drawing straight lines through chaotic dots.”
He cites the Challenger disaster: engineers had flagged the O-ring problem long before the shuttle exploded, but because launches hadn’t failed yet, managers assumed the system was sound. “The mistake was judging outcomes instead of process,” Brian explains. “Success can mask fragility.”
Georg noted the tension executives face: “Mature leaders log individual data points without over-extrapolating. Less mature ones take anecdotes as universal truths. The challenge is balancing storytelling, which organizations need, with the discipline to recognize when the data is just noise.”
When luck meets preparation
Not all randomness is destructive. Some of it, Brian notes, is opportunity. Many of the most transformative discoveries, careers, and companies began with accidents, missteps, or unexpected encounters.
“Serendipity can’t be planned,” Brian explains. “But it can be cultivated.”
One vivid example: during a London tube strike, commuters were forced to find new routes. Researchers later found that 5% permanently stuck with their new commutes because the disruption revealed a better option. “That’s forced experimentation,” Brian says. “Sometimes you only discover better paths when you’re nudged out of the old ones.”
Georg connects it to business: “If you give teams room to test, you get more shots on goal. Many will miss, but some hit… and the hits can be transformative.”
Trust as the currency of resilience
Randomness breeds anxiety. Brian’s work shows that trust is the differentiator in whether societies,and organizations, treat randomness as threat or opportunity.
“In high-trust societies, people assume mistakes are human, not malicious,” he explains. “They stick together in crises. In low-trust societies, the same shock sparks conspiracy theories and unrest.”
The analogy for business is direct. Trust between leaders and employees determines whether a team freezes during turbulence or leans in. Trust with customers determines whether a brand survives a stumble or is abandoned.
“Trust isn’t a soft factor,” Brian insists. “It’s the buffer that determines whether randomness breaks you or builds you.”
Vulnerability as strength
One of Brian’s most provocative arguments is that traditional ideas of leadership strength are counterproductive. The expectation that leaders must always project certainty, he argues, sets them up for failure.
“Pretending you know the unknowable is a recipe for disaster,” he says. “When leaders admit uncertainty, share their reasoning, and involve their teams, they don’t weaken authority. They build credibility. People respect honesty more than false confidence.” – Brian Klaas
Georg agrees. “In my experience, giving people the full, messy context actually strengthens alignment. Vulnerability isn’t weakness, it’s the foundation of trust.”
Designing resilient organizations
From these themes, a set of principles emerged that executives can act on:
- Accept uncertainty: Perfect prediction is impossible. Stop pretending otherwise.
- Invest in systems: Build structures that bend without breaking.
- Cultivate trust: It’s the shock absorber of resilience.
- Encourage serendipity: Innovation thrives where chance is allowed to land.
- Lead with vulnerability: Transparency builds credibility.
Georg summed it up crisply: “Leaders who succeed are those who normalize change. They prepare their organizations for turbulence — not stability — as the baseline.”
AI: enabling agility, not certainty
The conversation turned to artificial intelligence, where hype often masks the same illusion of control Brian warns against.
He draws a distinction: AI deployed in closed systems — like reading X-rays or optimizing energy grids — is relatively safe. But in open systems, where AI interacts with humans or with other AIs, unpredictability multiplies.
“AI won’t repeal chaos,” Brian cautions. “It’s subject to randomness itself. Leaders who expect AI to eliminate uncertainty are setting themselves up for failure.”
Georg reframes the point with an industry lens: “In language technology, or any business where humans and AI interact, you’re throwing pebbles into a pond already linked to other ponds by canals. Complexity multiplies. Compute power may increase, but physics says unpredictability doesn’t disappear. AI can amplify creativity and efficiency, but it won’t cancel chaos.”
The right mindset, Brian argues, is to treat AI as an amplifier: a tool that scales creativity, experimentation, and resilience.
“Use AI to make your organization more agile,” he advises, “not to chase the mirage of certainty.” – Brian Klaas
Closing reflections
Brian ends with a sobering but hopeful reminder. “Most of the future is unknowable. But that doesn’t make it unmanageable. Leaders who accept randomness and build for agility are far more likely to thrive when the unexpected arrives.”
Jasoncloses the discussion: “So the real job of leadership isn’t to eliminate chaos. It’s to build organizations that can dance with it.”
And Georg offers the final word for the C-suite: “Trust, resilience, agility — those aren’t slogans. They’re the foundations of leadership.”
Hear more
If you’ve enjoyed this look at the highlights from the latest episode of In other words, you can hear the entire episode now on our website, or subscribe via Spotify, Apple, or wherever you listen to your favorite podcasts.
