Over the past few weeks, I have had the opportunity to speak with several leaders about how organizations continue to innovate in increasingly unstable and technical environments.
I sincerely thank those who agreed to share their time, questions and, in some cases, concerns. These conversations were not academic discussions. They dealt with difficult decisions, exhausted teams, aging systems, weakened business models and artificial intelligence advancing faster than the organizations expected to integrate it.
This article continues those exchanges. It draws on their spirit and lessons to open a broader discussion about a topic that has become essential: innovation in a world where complexity often grows faster than our capacity to understand it.
By presenting innovation primarily as a technological promise, many organizations overlook what it actually exposes: vulnerabilities, blind spots, resistance and the ability to adapt when established reference points stop working.
Innovation is rarely the official subject.
It usually appears as a consequence.
A consequence of pressure. Complexity. A tightening market. A weakening business model. Aging infrastructure. Or an organization beginning to sense that yesterday’s answers will not be sufficient tomorrow.
One uncomfortable question follows: what is the organization still refusing to examine even though the signals are already present?
Innovation begins when certainty collapses
The OECD defines innovation through the effective implementation of a new or significantly improved product or process. Ideas, creativity and intentions are insufficient on their own. Something must be introduced to the market or brought into use within the organization.
This definition considerably expands the territory of innovation. A new way of organizing work, simplifying operations or making decisions can be an innovation in its own right. (OECD)
This distinction matters.
Many companies still confuse innovation with the accumulation of technologies. They launch laboratories, organize hackathons, purchase artificial-intelligence licenses and multiply proofs of concept.
Meanwhile, important decisions remain trapped in endless approval chains. Data stays fragmented. Teams manually copy information from one system to another. Responsibilities overlap. New tools are added without replacing the existing ones.
The shop window changes.
The underlying organization remains still.
Innovation then acts like a photographic developer. It makes visible what was already there: unnecessary processes, legacy systems nobody fully understands, ambiguous responsibilities, postponed decisions and comfort zones protected by habit.
Your organization may be accumulating debt that appears on no balance sheet
Companies understand financial debt and are beginning to recognize technical debt. Another form of liability is growing in the background: organizational debt.
Recent research describes it as the accumulation of outdated structures, policies and processes that limit an organization’s ability to adapt. It may result from poorly managed change, siloed operations, delayed decisions or compromises that gradually become permanent. Its consequences include lower productivity, reduced agility and weaker innovation capacity. (National Library of Medicine)
This concept helps explain why successful organizations can suddenly become vulnerable.
The problem does not always come from a lack of talent or resources. It may come from a long period of success that removed the need for deep self-examination.
Past success provides reassurance.
It legitimizes habits.
It transforms historical choices into supposedly permanent truths.
A process designed fifteen years ago survives because it once worked. An IT architecture becomes untouchable because the people who understood it have left. A commercial model remains in place because it still funds quarterly results. A management layer survives because removing it would open an uncomfortable discussion about accountability.
Each compromise appears reasonable when considered separately.
Their accumulation gradually creates an organization nobody would deliberately design today.
In my book, I examine this dynamic in Chapter 11, dedicated to organizations and structures. Innovation capacity is directly influenced by how a company distributes authority, organizes decisions, protects experimentation and circulates information.
AI accelerates production, but not necessarily understanding
Artificial intelligence currently acts as a powerful amplifier.
It speeds up research, automates processing, generates summaries, produces reports and allows small teams to reach an operational capacity previously reserved for much larger organizations.
The gains can be considerable.
Yet a coherent answer remains different from a correct one.
A polished report can rest on weak assumptions. A fluent summary can hide missing information. An efficient automation can execute a poorly designed process more quickly. A statistically plausible recommendation can become dangerous when applied outside its intended context.
The NIST AI Risk Management Framework emphasizes the need to measure system validity, reliability and robustness, and to continue testing and monitoring systems after deployment. It also recognizes the need for human intervention when an AI system cannot identify or correct its own errors. (NIST)
The European Union’s AI regulation also places human oversight among the structural principles applied through its risk-based approach. (EUR-Lex)
These regulatory and methodological requirements reflect an operational reality: computing speed does not replace judgment, responsibility or contextual understanding.
As tools become easier to access, human discernment becomes more critical.
A growing risk lies in the confusion between genuine competence and the appearance of competence. Someone can now produce in minutes a document that previously required several days. Its form creates an impression of mastery. Yet the person delivering it may not understand the reasoning, the underlying data or the consequences of the recommendations.
An organization can therefore automate its ignorance.
It can industrialize an error in judgment.
It can accelerate a process nobody has taken the time to challenge.
In my book, Chapter 14, which addresses the application of artificial intelligence, begins with vision and strategy. Before selecting tools, leaders need to determine what AI should change, what it should not decide and which outcomes the organization is pursuing.
“What can we automate?” is asked too early.
A more useful formulation is: “What value are we trying to create, and which part of this process still requires human understanding?”
The status quo resembles caution until the invoice arrives
Many organizations know they need to evolve, yet continue postponing important decisions.
The reasons are understandable: pressure on results, regulatory constraints, workforce fatigue, limited skills, economic uncertainty and competition between priorities.
The easiest subjects to postpone frequently involve invisible infrastructure: data quality, IT architecture, documentation, cybersecurity, technical governance, internal processes and skills development.
These subjects attract little attention while operations continue to function.
They become urgent after a failure, a data breach, the departure of a key employee, a regulatory inspection or the arrival of a more agile competitor.
The status quo operates like borrowed money.
The organization avoids the political, financial and human cost of transformation today. Tomorrow, it pays interest through delays, dependency, reduced competitiveness and exhausted teams.
Obsolescence rarely arrives overnight. It progresses through daily practices: a manual operation that remains tolerated, an exception that becomes the rule, a spreadsheet transformed into a critical system or a committee created to compensate for another committee’s inefficiency.
By the time the problem appears in financial results, resolving it usually requires more time, more money and more energy.
Simplification requires more intelligence than complication
The strongest organizations are not defined only by their sophistication.
They know how to make things understandable.
They clarify responsibilities. They reduce unnecessary interfaces. They remove meetings. They limit indicators. They document decisions. They select tools that genuinely replace previous tools. They help teams understand how their work contributes to strategy.
This simplicity requires a deep understanding of the business.
Adding a rule can quickly solve a local issue.
Removing ten rules requires understanding why they exist, what their removal will affect and how exceptions should be handled.
Complexity can look impressive.
Simplicity forces choices.
It exposes priorities, makes responsibility visible and makes inconsistencies harder to hide.
Many organizations spent decades pursuing maximum growth, diversification and permanent expansion. Several leaders I recently spoke with now talk more about visibility, recurring revenue, resilience and the ability to endure.
This shift changes how innovation is considered.
The objective is no longer limited to acceleration.
It also involves building systems that can evolve without continually exhausting the people who operate them.
Exploiting the present while exploring the future
Management research uses the term organizational ambidexterity to describe a company’s ability to exploit existing activities while exploring new technologies, markets and business models.
Exploitation prioritizes efficiency, control and incremental improvement.
Exploration requires autonomy, flexibility and experimentation.
These two approaches rely on different skills, metrics and sometimes different cultures. A sustainable organization must nevertheless learn to make them coexist. (Academy of Management Perspectives)
A company focused entirely on exploitation progressively optimizes what it already knows how to do. It becomes efficient, predictable and profitable until its market changes.
A company focused entirely on exploration may generate numerous ideas and experiments, but lack operational discipline, customers or revenue.
Leadership has to maintain this productive tension.
Protect current results without sacrificing future options.
Fund exploration without turning innovation into corporate theatre.
Accept that exploratory projects require evaluation criteria different from those used for a mature business.
This ambidexterity is a central theme in Chapter 8 of my book, which addresses vision, mission and strategy. A useful vision does more than describe a destination. It helps arbitrate between immediate pressures and the construction of the future.
Leadership becomes a discipline of clarity
For a long time, leaders were portrayed as the people who knew.
That posture is increasingly fragile in environments where technology, regulation and business models evolve simultaneously.
Leadership now depends less on producing an immediate answer and more on asking the questions that stop the organization from being collectively wrong.
Which weak signals are we ignoring?
Which assumptions remain untested?
Which essential capability depends on one person?
Which process are we automating without understanding its purpose?
Which decision are we postponing because it would challenge our previous choices?
Which information no longer reaches the executive committee?
This posture requires intellectual humility.
It also requires an environment in which an employee can report a risk, challenge an assumption or acknowledge an error without fearing a disproportionate sanction. Chapter 9 of my book examines how culture and psychological safety influence an organization’s capacity to see and discuss uncomfortable realities.
The World Economic Forum’s Future of Jobs Report 2025 finds that AI and data skills are gaining importance, while leadership, social influence, resilience, flexibility, creativity and curiosity are also becoming more valuable. Technological acceleration is therefore increasing the importance of the human skills required to interpret, decide and adapt. (World Economic Forum)
Protecting mental space becomes a strategic decision
Organizations can possess data, technology, budgets and talent, yet still fail to think beyond immediate urgency.
A succession of meetings, notifications, decisions and local crises gradually consumes all available mental space.
Teams spend their days keeping systems operational.
Managers absorb tensions.
Executives navigate between financial objectives, board expectations, customer demands and the transformations they need to prepare.
Innovation is eventually pushed into an abstract future.
An organization deprived of time for reflection becomes dependent on its reflexes. It repeats its previous actions more quickly. It treats symptoms without examining causes. It selects the most accessible solutions rather than the most relevant ones.
Protecting time to understand, learn and challenge assumptions is therefore not an intellectual luxury.
It is decision-making infrastructure.
Clarity comes before technology
The conversations of recent weeks converge around a simple but demanding idea: innovation begins with clarity.
Clarity about what must evolve.
Clarity about what is becoming fragile.
Clarity about accumulated compromises.
Clarity about the limitations of artificial intelligence.
Clarity about the real capacity of teams to absorb another transformation.
The organizations most likely to endure through the coming years will probably not be those that collect the greatest number of tools.
They will distinguish what matters from noise, simplify before circumstances force them to do so, protect experimentation and preserve enough human intelligence to interpret what machines produce.
Innovation is no longer limited to imagining the future.
It also means preventing the complexity of the present from suffocating the organization before it can reach that future.
