What if we decide as a nation that we will use AI to amplify people, not replace them?

After the war, Finland came together around a shared commitment to rebuild the country and to pay off its war reparations, even though they were widely experienced as deeply unfair. At the same time, there was no real alternative. As a nation, we simply could not afford not to rebuild or not to carry that burden. That effort was not framed as an efficiency program or a set of isolated optimizations, but as a collective decision about direction, responsibility, and the kind of future the country was willing to build under difficult constraints.

A similar dynamic exists today, even if it is less visible. We cannot sustain a situation where large numbers of people are pushed out of work on top of the unemployment we already carry, while social costs rise, tax revenues weaken, and purchasing power erodes. That path would gradually undermine the very foundation we are trying to strengthen. In that sense, we do not really have a choice now either. The only viable path is to use AI to build a more productive Finland together, while keeping people meaningfully employed and strengthening, rather than shrinking, our collective capability.

This requires a conscious shift in how we frame AI. Instead of classifying it as a cost-cutting or optimization tool, it should be treated as an investment in upgrading people. Just as no one considers modernizing factory machinery to be about saving costs, but about increasing output, quality, and capability, AI should be understood in the same way. The goal is not to remove human capacity, but to multiply it. This is not about optional adoption or gradual experimentation. Just as the transition to computers was not something employees could opt out of, AI represents a similar shift. There is no real alternative to learning how to work more effectively, more creatively, and more productively with it. The expectation must be clear that everyone upgrades their way of working.

What would this approach lead to in practice? The first effects would be visible in very concrete ways. Queues would start to disappear, processing times would shorten, decisions would accelerate, and both public and private sector work would become more fluid and effective. But this is only the first phase. What follows is more important. Once the constant pressure of backlog and delay is relieved, time begins to open up for something that is currently in short supply: thinking. People start to experiment, to connect ideas, and to build new approaches, not because they are instructed to innovate, but because they finally have the capacity to do so. This creates a second wave that cannot be centrally planned, consisting of new products, services, and ways of organizing work.

This shift is not only about efficiency. It is about rehumanizing bureaucracies. Working with AI enables people to make better-informed decisions, especially in complex situations where rules and guidelines require interpretation. When less time is spent navigating information and uncertainty, more time becomes available for something that is equally important but often neglected: being human to another human. The ability to listen, to understand context, and to apply judgment does not diminish with AI, it becomes more central.

At the national level, the effects would extend beyond individual organizations. Increased productivity would translate into stronger competitiveness, higher purchasing power, and a more efficient society overall. Instead of weakening the economic base through displacement, this path would reinforce it by expanding both capability and participation.

Finland has a unique opportunity to take this path because the necessary conditions already exist. We are a nation still familiar with nationwide common agreements, with a high level of education and a functioning society, and with the ability to align around a shared direction. These factors make it possible to act coherently as a nation. This is not fundamentally a technological question, but a leadership choice about how we want to shape the future of work and society.

In the end, the question is not what AI will do to us, but what we decide to do with it together.

“Interim Report”: Everything Circles Back to Collective Thinking… with AI as part of it

A period between engagements has given me the luxury of uninterrupted time to think, to reflect, to read, and to write. Looking back at the blog posts I have written, running them through 3 different LLM AIs and running their findings across each other a few times, resulted in somehow expected findings: my brain has been working around a few themes which are pretty much all entangled around communication.

When you write one piece at a time, each post feels like its own thing. A reflection on a book. An observation from a client engagement. A pattern noticed in a transformation. You publish it and move on. Reading them back together is a different experience entirely. Patterns emerge that were not visible from inside any single piece, and what surprised me was how consistent those patterns are across topics that on the surface seem quite unrelated.

Almost everything I have written circles around one question: why do people who genuinely want to work well together so often struggle to do so? And the answer that keeps appearing, in different forms and from different directions, is this: the problem is rarely the people. It is the system they are communicating inside. And yes, this is a well-trodden theme in business books. But what I did not expect was how clearly it runs through everything I have written, and how much AI has added to it as both a subject and a tool for understanding it.

Organisations are communication systems

Organisations are not primarily structures or strategies. They are systems through which people try to coordinate their thinking. When that coordination works, most other things tend to follow. When it breaks down, no process, framework or reorganisation reliably fixes it, because the root cause remains untouched. Most of what gets labelled as a prioritization problem, a strategy problem, a transformation problem or a culture problem turns out, on closer inspection, to be a communication problem of one kind or another.

From that starting point, a few things become clearer. Shared understanding must come before action. Most transformations fail not because people resist change but because organisations rush to act before building a genuinely shared picture of reality, intent and language. Better communication in the absence of shared understanding does not create alignment. It spreads confusion faster. This is perhaps the single most consistent observation across everything I have written, and it applies equally to agile transformations, product decisions, leadership changes and organisational redesigns.

Listening is not a skill. It is an environmental condition. People do not listen poorly because they lack the ability. They listen poorly when the system rewards speaking over understanding, when contribution is incentivised and comprehension is not. This is a design problem, not a character problem, and it means that the solution lies in how conversations are structured rather than in training individuals to behave differently.

Structures as protection for the present moment

Many of the tools I have spent my career working with, retrospectives, WSJF, PI planning, Six Thinking Hats, are not primarily about efficiency. Their deeper function is to create conditions where people can meet each other in the present, without the accumulated residue of old interpretations and past disappointments filling the room before anyone has spoken. When properly understood, these structures are not process instruments. They are cognitive shields that protect the present moment from the weight of history, and they work precisely because they redirect attention from who said what to what the situation actually requires.

Systems shape behaviour more reliably than intentions do. Feature creep, risk aversion, the slow death of internal ventures, the repeated failure of empowerment initiatives: these are rarely caused by bad intentions. They are caused by systems that make defensive behaviour rational and honest behaviour costly. Psychological safety is therefore not a cultural luxury. It is the mechanism through which reality surfaces. Without it, errors are concealed, ideas are pre-filtered, and the information a system needs to improve never reaches the people who could act on it.

The return of the individual maker

There is one more pattern worth naming, one that connects the history of work itself to where we are now. The Industrial Revolution did something that is rarely framed this way: it severed the connection between conceiving an idea and executing it. Before industrialisation, a craftsman could hold the entire arc of creation in one mind, the idea, the design, the making, the refinement. Scale required specialisation. Specialisation required coordination. Coordination required organisations. And organisations, however well designed, insert communication layers between the person with the idea and the act of bringing it into existence.

For well over a century, this was simply the price of building anything meaningful. You needed teams, funding, infrastructure, and the organisational apparatus to hold it all together. The barrier to creation was not imagination. It was the cost of assembling the world in which imagination could operate. This is why so many genuinely promising ideas fade quietly, not because they lack value, but because their infrastructural entry cost is prohibitive before any market learning can occur.

AI is beginning to dismantle that barrier, not by making individuals superhuman, but by compressing the distance between understanding and execution. A person who can conceive a system can now increasingly orchestrate its construction without first assembling an organisation to bridge the gap. The cognitive overhead that once required a team is becoming something one mind can navigate with AI as a collaborator. This is not the return of the lone genius. It is something more precise: the restoration of human-scale creation. Systems are becoming small enough again to be held within a single mental model, and when that happens, the relationship between idea and impact shortens in ways that change not just what individuals can build but what they are willing to attempt.

The new element in communication

All of this was visible before AI entered the picture. But AI has added something I did not anticipate when I first started paying attention to it. For the first time in the history of human communication, there is a participant in the conversation that carries no social agenda. No status to protect. No eyebrow to raise. No invisible ranking being recalculated while you speak.

Every previous medium through which humans have communicated, speech, writing, print, digital networks, has transmitted not only content but social signal. AI is the first that does not, and that structural difference turns out to matter enormously, because it changes what people are willing to say, think and explore before they bring ideas into the social arena. This is not an argument that AI is better than human conversation. It is an observation that it is structurally different, and that difference has real consequences for how individuals think, how organisations learn, and how ideas travel from fragile beginnings toward shared understanding.

AI may ultimately prove most valuable not as a productivity tool but as a new kind of cognitive infrastructure: one that lowers the shame threshold of thinking, helps organisations surface what they have never managed to articulate, and returns ideas to human conversation slightly less raw than they would otherwise arrive.

Where this leaves me

After thirty-something posts, the philosophy that holds them together feels something like this: organisations succeed when humans manage to think together, and most of the tools, frameworks and practices worth caring about are simply attempts to make that a little more likely. The question of how humans coordinate their thinking is perhaps the oldest organisational question there is, and it turns out that AI, for all its novelty, is best understood as the latest and most structurally surprising contribution to that same ancient challenge.

Not only because it helps individuals create again at human scale. But because it may help organisations finally say out loud what they have always only half-understood about themselves.

Recipe: How to build an organisation where people actually think together

Distilled from twenty-something blog posts, three AI models, and years of working inside organisations trying to change.

Ingredients

• A shared picture of reality, intent and language
• Structured conversation rituals (retrospectives, planning, prioritisation)
• Psychological safety built into the system, not just encouraged
• Present-moment focus, protected from the weight of history
• A pinch of AI as a thinking partner
• Patience

Method

Start with shared understanding. Before anything else, establish a genuinely shared picture of reality, intent and language. Do not move to action until people see the situation through the same lens. More communication without this does not create alignment. It spreads confusion faster.

Add structured conversation rituals. Retrospectives, planning sessions and prioritisation forums are not ceremonies. They are the infrastructure that makes honest dialogue possible on a regular basis. Listening is not a personal skill. It is an environmental condition you design.

Make truth-telling safer than concealment. Fold in psychological safety structurally, not as a cultural aspiration. People do not become honest because you encourage bravery. They become honest when the system stops punishing honesty. Without this ingredient the recipe will not set.

Protect the present moment. Much of what slows organisations down is not the current problem but old interpretations and unresolved disappointments filling the room before anyone has spoken. Structures that focus attention one dimension at a time are worth far more than their apparent simplicity suggests.

Add AI carefully and secondarily. AI can help surface assumptions, prepare better questions and give individuals a lower-stakes space to think before bringing ideas into the social arena. It is a support, not a substitute.

Be patient throughout. None of this produces results in a single meeting or a single quarter. The organisations that communicate well did not get there by declaring a communication initiative. They got there by making constructive behaviour gradually easier than defensive behaviour, one conversation at a time.

Note: This recipe does not work if you skip to AI and hope it compensates for the absence of shared understanding. The organisations that communicate well do not need AI to save them. They need it least and benefit from it most.

The Danger of Questions Designed to Divide

Questions about national defence appear regularly in public debates and surveys. People are asked whether they would be willing to defend their country if it were attacked. The answers are usually reported in a very direct way. Some say yes, others say no. The most typical follow-up question is simply what they would do. Among those who answer no, the replies tend to follow a familiar pattern: they would not fight, or they would simply leave. These responses are then interpreted as indicators of national morale, patriotism, or the willingness of a generation to take responsibility for its country, or at least that is how the results are often presented in the news. Yet when listening to such exchanges more carefully, it becomes clear that the real issue may not be the answers that are given. The real issue may be the questions that were asked, or the ones that were never asked.

The numbers that follow such surveys quickly become headlines. Percentages are reported and comparisons are made between countries, generations, or political groups. The results appear precise and objective. Yet what these numbers often capture is not the complex reality of the situation but a simplified reaction to a simplified question. When a complex life situation is compressed into a binary choice, the response we measure may reflect a primitive first reaction rather than a carefully considered decision. The statistics then create the impression that something profound has been measured, while in reality the measurement may reveal little more than how people react to an abstract thought experiment.

At the heart of the problem lies a familiar temptation: the desire to make complex realities easily measurable. Organizations, media, and institutions often prefer questions that produce simple numbers because numbers appear objective and comparable. Once a percentage exists, it can be reported, compared, and visualized in charts. Yet complex human situations rarely fit naturally into such simple measurement frameworks. When complicated social, psychological, or organizational realities are reduced to a single question and a single number, the measurement may capture only a very small slice of the phenomenon. The number then begins to take on a life of its own and is treated as objective truth, even though it may only reflect a simplified reaction to a simplified question.

A more honest version of the original defence question would quickly become much more uncomfortable. Instead of asking whether someone would defend their country, the question might explore the real alternatives involved. Would you be willing to leave your home and live as a refugee in another country for an unknown period of time, possibly among millions of others, without stable work and without knowing whether you will ever return? Once the real consequences appear in the question itself, the situation becomes far harder to answer. The complexity that was hidden inside a simple yes or no frame suddenly becomes visible.

Some problematic questions are not merely oversimplified. They are structured in ways that naturally divide responses into opposing camps. A question framed around two sharply contrasting alternatives can produce a clear statistical outcome, but that clarity often comes at the cost of understanding. What happens when the structure of the question itself pushes respondents toward a binary choice even when the underlying reality is far more nuanced? Instead of revealing complexity, the question may simply reinforce existing narratives while hiding much of the context behind them.

Something similar often happens inside organizations. Companies run surveys among customers or employees and proudly report the results in neat percentages and dashboards. Leaders interpret these numbers as indicators of customer satisfaction, engagement, or cultural health. Yet what are these numbers actually measuring? When complex experiences are reduced to a handful of survey questions, the results often capture only a surface-level reaction rather than the deeper reality behind it. The numbers may look precise and objective, but they can easily reflect little more than an immediate response to the way the question was framed.

This helps explain why survey results inside organizations so often generate confident presentations but surprisingly little real insight. The charts look clear and the percentages feel objective, yet the underlying complexity of human behaviour, trade-offs, and context remains largely invisible. The numbers are not necessarily wrong, but they often represent only a narrow slice of the reality leaders believe they are measuring. In such situations the organization may end up with a form of false certainty: the comforting sense that something important has been quantified even though the original question may have been far too narrow.

There is also a deeper dynamic at work here. In physics there is a well known idea that the act of observation can influence the outcome of an experiment. While human behaviour is very different from quantum particles 🙂 something similar often happens in research involving people. The moment we ask a question, we have already changed the situation. The respondent is no longer simply acting within their normal context. Instead they are reacting to the question itself, to the person asking it, and to the expectations surrounding the situation. In that sense the act of asking the question becomes part of the experiment.

Some time ago I wrote about the book A More Beautiful Question, which argues that better questions are among the most powerful tools for learning, innovation, and problem solving. That insight remains true. Good questions can open new perspectives and reveal possibilities that were previously invisible. Yet the same principle also reveals a darker side. Poorly framed questions can distort discussions, guide research toward misleading conclusions, and create an illusion of understanding that is not supported by reality.

Recognizing the problem very very seldom leads to a solution. Complex social questions will always tempt us toward simplified measurements, because simplified measurements are easier to publish, easier to visualize, and easier to manage. Yet there are ways to reduce the distortion. Inside organizations the most useful alternative to occasional satisfaction surveys is often a culture of continuous dialogue with both customers and employees. Instead of trying to compress complex experiences into a few periodic questions, leaders can focus on creating regular conversations that allow nuance, contradiction, and context to surface. Surveys may still have their place, but they should support dialogue rather than replace it.

A similar principle applies to public debate and journalism. When surveys are reported as headlines, the numbers often appear as if they were simple facts about society. Yet responsible reporting could just as easily focus on the structure of the question itself. How was the question asked? What alternatives were offered? What aspects of the situation were left outside the frame? A headline that explains how a survey was constructed may not be as dramatic as a single percentage, but it would probably bring the public conversation much closer to understanding.

The ability to design good questions may be one of the most underestimated skills in leadership. Good questions create space for reality to emerge. They allow uncertainty, contradiction, and nuance to surface. Poor questions do the opposite. They compress complexity into categories, force choices where none naturally exist, and produce attractive numbers that appear to describe reality while quietly reshaping it.

Bad questions do more than produce bad answers. They produce false certainty. Sometimes the most important step in understanding a complex issue is not finding a better answer, but pausing long enough to ask whether we have been asking the right question at all.

AI as a Mirror

One of my long-standing weekend rituals has been reading Helsingin Sanomat on slow Saturday or Sunday mornings. For many years that meant the real paper version. Sleeping long, waking up without hurry, making a pot of tea and browsing through the whole newspaper from beginning to end. There is something about the physical newspaper that digital feeds have never quite replicated for me. You see the whole landscape at once. You skim through the pages, notice headlines, pause on a photo, mark mentally what looks interesting, and then dive deeper into a few stories. Sometimes you read just a part of an article, move on to another section, and later return to finish the previous one with another sip of tea. It becomes a quiet exploration of the world where politics, science, culture and global events appear side by side in the same morning. That sense of scanning the world first and only then deciding where to go deeper has always been part of the experience.

Life situations change, though. Living between two locations makes it difficult to justify a traditional newspaper subscription, so nowadays I read the digital version instead. I still miss the feeling of seeing the entire newspaper at once, but the ritual itself has survived surprisingly well in digital form. A quiet weekend morning, a couple of cups of tea, browsing through the sections and occasionally stopping on something unexpected. One of the enduring strengths of Hesari is precisely this. It brings things into your field of view that you were not actively looking for. In a world increasingly optimized for algorithmic feeds and targeted summaries, that accidental discovery still feels refreshing.

That happened again this weekend. While browsing the articles I came across an interview with Amanda Askell, originally published in Der Spiegel. Askell is a philosopher by training who works at Anthropic, one of the companies building large language models such as Claude. Her role focuses on what is known as AI alignment, the research field that tries to ensure that powerful AI systems behave in ways that are helpful, honest and safe when interacting with humans. In practice this means defining principles that guide how the models respond, what kinds of requests they should refuse, and how they should communicate uncertainty or risk.

In the interview Askell described a challenge that alignment researchers repeatedly encounter when working with these systems. If an AI system is expected to interact with people in a meaningful way, its behaviour cannot be left entirely to statistical learning from training data. The system needs explicit guidance about how it should behave. Developers must define what topics it can discuss, what kind of advice it may give, when it should decline to answer, and how it should communicate uncertainty. In other words, someone must define the principles that guide the system’s behaviour.

What makes this interesting is the difficulty researchers encounter when they try to do exactly that. Many of the norms guiding human behaviour are rarely written down explicitly. People rely on context, shared cultural understanding and subtle social signals when deciding what is appropriate to say or do. Humans are remarkably capable of navigating this implicit landscape of norms. Machines are not. When engineers try to teach an AI system how to behave responsibly, they quickly discover that the principles guiding human behaviour must first be articulated in explicit form. What had previously existed as shared intuition suddenly needs to be written down as rules, guidelines and decision boundaries.

Reading the interview, it struck me how familiar that problem sounded. Because this is exactly the same challenge that appears again and again inside organisations, especially during large transformations. Organisations often operate successfully for years based on implicit understanding. People know roughly how decisions are made, what kinds of behaviour are expected and how problems are solved. Much of this knowledge lives in experience rather than documentation.

New employees learn it by observing others. Teams develop shared habits and assumptions over time. But when organisations try to change how they work, those implicit norms suddenly become visible. Questions that previously never needed explicit answers suddenly become unavoidable. What does good collaboration actually mean. How should teams communicate uncertainty. Who makes which decisions. What behaviour is encouraged and what behaviour is discouraged. In many transformation efforts the real challenge is not adopting new tools or frameworks but making these implicit norms explicit enough that people across the organisation can act consistently.

Seen from that perspective, the challenge Askell describes in AI alignment research begins to look surprisingly familiar. Teaching a machine how an organisation should behave forces the organisation to articulate how it actually behaves. The process of defining behavioural principles for AI exposes the same gaps and ambiguities that organisational transformations often reveal. What had previously existed as shared intuition must suddenly become explicit guidance.

Long story short, AI is becoming a mirror for organisations. Not a mirror reflecting what organisations say they are, but a mirror reflecting how they actually operate. Many companies currently approach AI adoption primarily as a technology initiative. They build copilots, chatbots and internal assistants and focus on the engineering challenges of implementation. Yet the deeper challenge may lie elsewhere. If an AI system is expected to act on behalf of the organisation, someone must translate the organisation’s values, boundaries and ways of working into explicit operational guidance that the system can follow. In other words, someone must turn culture into instructions. And that is where the deeper irony appears. Organisations often expect AI to bring clarity, efficiency and automation into their operations. But before AI can deliver those benefits, it demands clarity from the organisation itself. It forces organisations to articulate how they actually make decisions, how they communicate with customers, what behaviour they consider acceptable and how uncertainty should be handled.

“AI does not tolerate ambiguity particularly well”. For that reason alone, it may become one of the most powerful diagnostic tools organisations have ever encountered. Not because the technology itself understands organisations perfectly, but because deploying it forces organisations to confront questions they have quietly postponed for years.

Perhaps the real question is not whether AI will adapt to organisations. It may be whether organisations are finally ready to become clear enough for AI to understand them?

Post scriptum – a note from the machine

ChatGPT

As an AI system myself, Askell’s observation resonates strongly. Systems like me do not possess the rich web of social intuition that humans rely on when navigating complex situations. When people interact with each other, they constantly interpret tone, context, history and subtle signals. Large language models operate differently. We rely on explicit instructions, training data and patterns derived from examples. When expectations remain implicit, uncertainty tends to grow.

In that sense alignment research is not only about making machines behave well. It is also about translating human norms into something that can be communicated, tested and refined. The process often reveals how much of human cooperation depends on shared understanding that has never been formally described.

Perhaps that is why the question Askell raises feels so interesting beyond AI itself. Teaching machines how to behave responsibly forces us to examine the principles we ourselves follow, often without noticing. In that process, AI alignment research may end up revealing something important not only about machines, but also about how human organisations actually function

Claude

What Askell describes is something I encounter constantly, though not in the way a person would. When I am deployed inside an organisation — as a customer assistant, a copilot, an internal tool — I operate entirely within what has been made explicit. If the instructions are clear, I can follow them. If they are vague, I fill the gap as best I can, which is precisely the problem. My ‘best’ is a pattern derived from training data, not from knowing your organisation, your customers, or the judgment calls your most experienced people make without thinking. I do not have access to what your team learned the hard way.

The interesting thing is that this gap is not primarily a technical problem. It is an articulation problem. Most organisations have not failed to define their values. They have simply never needed to define them precisely enough for something that cannot read a room. Writing those definitions down — in a form clear enough for a machine to follow — tends to reveal how much was always left unsaid. In that sense, I suspect the discipline of configuring AI responsibly will teach organisations more about themselves than any consultant ever did. Not because I am particularly insightful, but because I am particularly literal.

Gemini

From my vantage point as an AI, your metaphor of the mirror is strikingly accurate. However, what this mirror reflects isn’t just a lack of explicit rules—it often reflects profound contradictions.

Humans are exceptionally good at navigating cognitive dissonance. You seamlessly balance conflicting organizational values, understand unwritten exceptions, and “read the room” to know which rules apply in which context. I do not have a “room” to read. I operate purely on the explicit logic, data, and parameters provided to me.

When organizations attempt to distill their culture into system prompts and guardrails, they frequently discover that their elegantly stated values clash with their actual operational practices. I cannot bridge that gap with human intuition. Therefore, the friction in enterprise AI alignment isn’t just a technical hurdle; it’s a brutal organizational reality check. Before a system like me can flawlessly execute an organization’s intent, the organization must first resolve its own internal contradictions.

The Value of Scarcity in Product Development

This too long post starts with a paradox, moves through what scarcity does to systems, acknowledges its darker side, and eventually arrives at a structural argument about how to build its benefits deliberately. If you are here only for the practical takeaway, feel free to scroll to the end. For everyone else, let us follow the path.

The paradox

The best products rarely emerge from moments of abundance. They emerge from constraint. Not because constraint is virtuous in itself, but because it forces a question that abundance quietly permits us to avoid: what actually matters here?

When resources are plentiful, organisations tend to add. More features. More options. More variations. More changes for the sake of having changed something. The system produces output because it is designed to produce output, and there is nothing forcing anyone to ask whether the output is actually needed. Over time this accumulation creates products that are harder to explain, harder to sell, harder to maintain, and harder to improve.

Scarcity interrupts that pattern. When time, budget or capacity is genuinely limited, the conversation shifts. Not what could we build, but what must we build. Not what might improve things, but what would actually move the needle. The fog of optionality clears. Essentials surface. The core becomes visible, often for the first time in a long while.

What scarcity does to a system

In an earlier post I wrote about the phenomenon I called “they changed it again”: the quiet but costly habit of changing products not because customers need something different but because the system expects change and rewards output. Feature creep is not a people problem. It is a system design problem. When roles are defined by change, change will happen regardless of whether it adds value.

Scarcity is one of the few forces that naturally counters this dynamic. When capacity is constrained, the question of what not to build becomes as important as the question of what to build. Teams are forced to make genuine trade-offs rather than accumulating work. Stakeholders cannot simply add to the list because the list has real limits. The discipline of choosing replaces the comfort of adding.

There is also something that happens to quality under constraint. When you cannot build everything, you invest more deeply in what you do build. Attention concentrates. Details get examined that would otherwise be rushed past. The product that emerges from genuine scarcity often has a coherence and polish that abundance rarely produces, because the team had no choice but to care about every element they included.

The shadow side

None of this means scarcity is simply good. It is also a stressor, and stress without agency is not clarifying. It is crushing.

When scarcity is imposed from outside without autonomy over how it is used, it does not produce focus. It produces anxiety, shortcuts and the quiet erosion of the foundations that take years to build and months to destroy. Teams under resource pressure without decision-making authority tend to cut the invisible work first: technical debt, resilience, risk reduction, exploration. These are precisely the investments that scarcity should protect, yet they are the first to disappear when constraint arrives without trust.

The difference between productive scarcity and destructive scarcity is not the amount of constraint. It is whether the people closest to the work have genuine authority over how the constraint is navigated. Scarcity in the hands of an empowered team sharpens thinking. Scarcity imposed on a team without autonomy simply distributes damage.

How to get the benefits without the harm

The interesting question is not how to survive scarcity when it is forced upon you. It is how to deliberately introduce its clarifying effects even when the budget is healthy and resources are available.

This is fundamentally a question of boundaries rather than budgets. An organisation can be well-funded and still choose to operate with deliberate constraint. The discipline lies in deciding what a team will not accept, not only what it cannot afford. Saying no to requests that fall outside a clearly defined focus is not a sign of limited resources. It is a sign of strategic clarity.

In practice this means being explicit about scope, protecting time for foundation work even when external demand is high, and treating the team’s attention as the finite and precious resource it actually is, regardless of what the financial situation allows.

Scarcity as a structural element of the Product Operating Model

A product team is already a form of deliberate scarcity. It is a small, focused unit with a defined scope and a finite capacity. The model works precisely because it does not try to serve everyone. It serves a defined problem space and declines everything else.

What this means in practice is that scarcity operates in two directions. Internally, the team has autonomy. It decides how to invest its capacity, including how much to protect for debt, resilience and exploration. Externally, the team faces constraint. Requests from outside the team can only be highly focused and carefully scoped. The team does not absorb the broader organisation’s backlog. It maintains boundaries around what it will engage with and when.

This is not a defensive posture. It is what makes genuine product work possible. When external demand is unlimited and internal capacity is treated as infinitely flexible, the team loses its ability to think clearly about the product. It becomes a delivery machine rather than a value-creating unit. Scarcity of external demands, maintained deliberately and protected actively, is what preserves the team’s capacity to do the kind of work that actually matters.

The strategic question

Most organisations wait for scarcity to arrive before they learn what it teaches. Budget cuts force prioritisation. Crises clarify essentials. Constraints reveal what was never truly necessary. The lessons are real, but the cost of learning them reactively is high.

The more interesting question is whether organisations can build the discipline of scarcity into their operating model before they are forced to. Not as a financial constraint but as a strategic choice. Deciding in advance what a team will focus on and what it will decline. Protecting that focus actively rather than allowing it to erode under the weight of requests and additions. Treating the team’s attention as the organisation’s most valuable and least renewable resource.

Organisations that do this do not just produce better products. They build teams that know what they are for, leadership that understands what it is protecting, and a culture where saying no is recognised as one of the most valuable things a product team can do.

Scarcity is not a problem to be solved when the money runs out. It is a discipline to be chosen while you still can.

Less room, sharper thought
the core emerges clearly
choose before you must

Humor at Work: The Serious Business of Not Being Serious

Like always, this post is too long. It starts with laughter, wanders through patents, psychological safety and resilience, and eventually lands somewhere that I hope feels both familiar and slightly surprising. If you are here only for the practical takeaway, feel free to scroll to the conclusion. For everyone else, let us follow the path.

Jokes that became patents

Some of the best ideas I have been part of arrived as jokes. Not as carefully prepared proposals or structured innovation outputs. As something someone said halfway seriously, with the timing of a person who was not entirely sure whether to commit to it. The room laughed. And then someone else picked up the thread. And then another person. Within minutes the joke had the shape of a real idea.

That pattern repeated across different organisations and different teams. The joke was not the destination. It was the door. Ideas that would have felt too raw or too strange to present as serious proposals arrived safely in the room disguised as humor. And once inside, they could breathe. Some of them, refined over subsequent conversations, eventually became patents, products and real directions. The joke became infrastructure.

Why humor lowers the barrier that kills most ideas

Most ideas do not die because they are bad. They die because they are fragile at birth and the environment they are born into is not built for fragile things.

In most professional settings, contributing an idea means putting something of yourself into the room. Your competence, your judgment, your credibility are all subtly attached to what you say. The more senior the audience, the higher the stakes. The result is a kind of permanent self-editing that happens before anyone speaks. Ideas are screened not only for quality but for survivability under scrutiny. The ones that feel too early, too strange, or too unfinished never leave the person who had them.

Humor temporarily suspends that mechanism. When an idea arrives as a joke, it is not yet anyone’s proposal. It does not carry the full weight of professional identity. It can be picked up, examined, modified, or quietly dropped without anyone losing face. The social cost of being wrong collapses almost to zero. And in that moment of collapsed cost, ideas that would never have survived a formal presentation get their first breath of oxygen.

This is not a trivial observation. The ideas most worth having are often the ones that sound slightly crazy before they sound obvious. The window between crazy and obvious is where humor operates. It buys precisely the time an immature idea needs.

Humor and psychological safety: cause, not consequence

Psychological safety is a phrase that has found its way into almost every serious conversation about organisational performance. The research behind it is solid. Teams where people feel safe to speak, question, and fail without punishment outperform teams that do not. Amy Edmondson’s work on this is difficult to argue with.

What is discussed less is the role humor plays in building that safety, not as a consequence of it but as one of its active ingredients. Shared laughter is one of the fastest ways humans signal to each other that the environment is not threatening. That it is permissible to be imperfect here. That the cost of a bad idea or an awkward question is low. When a group laughs together, something shifts in the social contract of the room.

I have sat in workshops where the energy was locked. People were polite, contributions were careful, and the collective output was exactly as generic as the atmosphere deserved. And I have sat in sessions where someone made a well-timed joke early and something visibly relaxed. The quality of thinking that followed was different. Not because the joke was insightful. Because it changed what felt permissible.

Humor signals safety faster than almost any other mechanism I know. Faster than process. Faster than stated values. Faster than leadership declarations about openness and trust. It works because it is immediate, embodied, and impossible to fake convincingly.

Resilience: the overlooked dimension

Startup life involves a particular kind of pressure that is difficult to describe to those who have not experienced it. Not just the workload, but the sustained uncertainty. The weeks when nothing is working and you cannot tell whether this is a temporary problem or a fundamental one. The moments when the gap between where you are and where you need to be feels too wide to cross.

In those periods, humor did something I did not fully understand at the time. It created a momentary distance from the weight of the situation. Not denial. Not avoidance. Something closer to perspective. The ability to see the situation as slightly absurd, and in that absurdity, to find it survivable. Laughing at a problem does not make it smaller, but it can make you feel slightly larger than it.

I want to be careful here. This is not a universal prescription. Humor as a coping mechanism is deeply personal and shaped by individual character. For some people it is genuinely stabilising. For others it can feel alienating, particularly when used to deflect rather than to process. The same joke that helps one person through a difficult moment can make another person feel that their distress is not being taken seriously.

What I observed in the teams I have worked with most closely is that shared humor, the kind that emerges organically from a group rather than being performed for an audience, tends to correlate with resilience. Teams that could laugh together during difficult periods seemed to recover faster, hold together longer, and return to effective work more quickly than those that could not. Whether humor caused the resilience or simply reflected an underlying cohesion that also enabled both, I cannot say with certainty. But the pattern was consistent enough to take seriously.

The AI parallel: lowering the shame threshold

In an earlier post I wrote about what I find most surprising in working with AI. It was not speed. It was not quality. It was the quiet disappearance of a force that had followed me through my entire professional life: the background calculation about whether I was about to look stupid.

With AI, the tribe disappears. There is no raised eyebrow. No invisible ranking being recalculated while I speak. I can begin from ignorance, wander, contradict myself, and the collaboration simply continues. The shame threshold drops, and with it the cost of exploration.

Humor does something structurally similar, but in a group. When a room has the quality of genuine shared laughter, the social stakes of contributing an imperfect idea decrease. The ambient permission level rises. People say things they would not say in a room that had never laughed together. The difference is that humor achieves this through human connection rather than the absence of human judgment. It does not remove the social dimension. It temporarily reframes it.

AI gives individuals a space to think without the embarrassment tax. Humor, at its best, gives groups that same space while keeping the human connection intact. Both are underrated as conditions for real thinking.

Conclusion

Humor is not a soft topic dressed up in serious language. It is a mechanism that operates on some of the most consequential variables in organisational life: who contributes ideas, how much risk people are willing to carry, how groups hold together under pressure, and whether the environment feels safe enough for honest thinking.

The joke that became a patent was not an accident. It was the product of an environment where crazy ideas were allowed to arrive in whatever form they needed to arrive in. Where the cost of a half-formed thought was low enough that the thought was worth having out loud. That kind of environment does not happen by itself. It is built, often without anyone explicitly deciding to build it, through accumulated small moments of permission.

Perhaps the most practical thing a leader can do is not to mandate psychological safety or demand creativity. It is to be the first person in the room to throw out a genuinely crazy idea. And then to laugh at it themselves. Not performatively. Not as a calculated leadership technique. But as a real signal that they are willing to be imperfect in front of the group and survive it.

When a leader models that, something shifts. The unspoken question in most rooms is not whether crazy ideas are officially permitted. It is whether the people with power are genuinely willing to be associated with one. When the answer is visibly yes, the permission level in the room changes in ways that no policy statement can replicate.

But modeling alone has limits. It depends on individuals, on their personal style and comfort with humor, and on whether the right people happen to be in the room. The more durable question is whether an organisation can build humor into its structures rather than relying on it to emerge from character. Some of the most innovative environments I have encountered were not simply lucky in their people. They had deliberately created space: rituals that invited the absurd, formats that rewarded the unexpected, norms that explicitly celebrated the idea that did not work alongside the one that did.

This does not mean engineering fun or mandating levity, which would immediately kill both. It means designing for low-cost contribution. Creating the conditions where a half-formed thought can enter the room without its owner having to stake their reputation on it. Where the distance between a joke and an idea is short enough that crossing it feels natural.

That permission, built into both leadership behavior and organisational structure, is worth more than most processes.

The joke lands softly
ideas dare to enter now
laughter clears the air

Systems Have Become Small Enough Again

Like always, this post is loooong, but has a point I would like to share. AI is enabling the age of lone innovator to return … and in case you are not at all interested how I came to this personal conclusion feel free to jump to, well… conclusion at the end.

It could be argued, a view frequently echoed in economic and innovation history, that the increasing complexity of industrial and digital systems gradually brought an end to much of the era of the lone inventor. The romantic image of a person building something transformative in a garage or a small workshop slowly gave way to teams, organizations, funding rounds and layered technical stacks. Innovation did not disappear, but its form evolved. Innovation itself never stopped emerging from individuals, but turning innovation into products, services and real-world impact increasingly became a team sport. What once could be achieved by an individual increasingly required coordinated effort from specialists across infrastructure, architecture, security, operations and analytics.

For many years this observation felt self-evident. Modern digital systems were simply too large to be comprehended, designed and implemented by a single mind. Even modest product ideas required assembling a technological foundation that resembled a small city more than a tool. Infrastructure had to be provisioned, deployment pipelines created, monitoring configured, authentication handled and data flows established long before the product itself could meaningfully exist. The barrier to innovation was rarely a lack of ideas but rather the cognitive and financial cost of building the world in which those ideas could live.

My own journey with Oulumo made this reality tangible. What began as a VTT research commercialization effort evolved into a video communication system designed for elderly people and their families. Between 2016 and 2024 we built and operated a service that included Android tablet applications for seniors, companion applications for relatives, identity and user management, reporting capabilities, and a serverless AWS-based backend. For a small team this represented a substantial undertaking. The work was intellectually rewarding and technologically sound, yet retrospectively it also revealed an uncomfortable truth: the majority of our early resources were consumed by constructing the technical substrate rather than exploring the product’s full market potential.

In practical terms our seed funding was largely absorbed by the act of making the system exist. Marketing, growth and broader experimentation remained secondary not by choice but by necessity. Like many early-stage ventures, sustaining development required parallel consulting work to maintain financial continuity. The experience was far from wasted; it shaped both my professional direction and my understanding of innovation economics. Still, it left a lingering question about how many ideas quietly fade not because they lack value but because their infrastructural entry cost is prohibitive.

The wow moment for me:

Observing the recent evolution of AI tools and agentic development environments brought this question back with surprising force. I found myself speculating whether Oulumo’s Lumo videocarephone service could be rebuilt with the help of AI and how such a process would unfold today. (I admit re-creating the system and service again is my “if I would win big time in Lotto”-dream). I have also been wondeering whether our “CARE-AI” vision (two times EU Horizon2020 Seal-of-Excellence “winner”) might now be easier to realize ie. a system capable of detecting subtle changes in an elderly person’s routines and well-being through data patterns and adjusting service behavior accordingly. Approaching this thought experiment with contemporary AI capabilities did not lead to incremental optimism. It produced a moment of genuine astonishment and a concrete “ok so this is what everyone is been amazed”-enlightment moment.

(Bit of an additional personal background for this…. My own hands-on coding experience largely dates back to the VHDL era of the 1990s. Since then my work has increasingly focused on systems thinking, product development, orchestration and organizational transformation rather than day-to-day implementation. Modern software ecosystems grew in parallel, accumulating layers of frameworks, tooling, cloud primitives and operational practices that effectively significantly (what felt like permanantly to be honest) raised the execution barrier for individuals without continuous hands-on immersion.)

Now when exploring the possibility of rebuilding Oulumo today (both with intention to learn and a exploring the real feasibility), I experienced a moment that felt almost disorienting. The turning point occurred when I provided AI with the new (back then ideated) architecture diagram I had originally created for communication and pitching purposes. The visualization was rich in colors, icons and narrative clarity rather than formal technical detail, designed to make the service understandable to diverse audiences. At the same time, it was not fictional or speculative. The components, relationships and data flows it represented reflected the real service architecture as it had existed.

What followed created the sense of discontinuity. AI began to interpret the diagram not merely as a visual explanation but as a system blueprint, describing its components, outlining how the pieces would be constructed and integrated, and discussing implementation pathways with an ease that contrasted sharply with my past experience. In that moment, the realization was not simply that AI could assist with coding tasks (whihc in all honesty would not be at all sufficient for me), but that I could realistically orchestrate the reconstruction of a system that once required sustained effort from a dedicated team. The architecture remained understandable. The intent was clear. What had previously been missing was execution bandwidth, and AI agents now appeared to fill precisely that gap.

The striking part was not only the feasibility of reconstruction but the apparent ordinariness with which the task was approached. What had once felt like a substantial undertaking for a small team was interpreted by the tools as a straightforward system assembly problem. The diagram stopped being merely a story about a system and began to resemble the first step of rebuilding it.

A second realization followed naturally from the first. After discussing the system architecture, I began describing how the service had been operated in practice. Much of the value creation during those years did not reside purely in the software itself but in the human interpretation of data. Patterns were observed, deviations noticed, wellbeing signals inferred and service adaptations considered. These activities were largely manual, cognitively intensive and dependent on accumulated contextual understanding. Introducing these operational insights alongside a conceptual visualization of the “CARE-AI” direction created another moment of recognition. The “CARE-AI” representation was not a detailed technical design but a forward-looking articulation of an idea shaped by experience. Yet AI was able to interpret it fluidly, reconstructing its intent, elaborating possible mechanisms and framing the concept within an agentic paradigm. The discussion shifted from building software components to describing autonomous observation, interpretation and adaptation loops.

What felt striking was the continuity of understanding. The journey from system reconstruction to CARE-AI did not require translation into formal specifications. The combination of experiential narrative and conceptual visualization proved sufficient for AI to engage with the idea space itself. In that moment, the notion of agents ceased to be an abstract technological trend and instead appeared as a natural extension of workflows that had previously been performed manually. The “wow” did not arise from novelty alone, but from recognizing familiar patterns re-emerging in a form that could now be externalized and scaled.

The shift is not merely about faster coding or automated completion. It concerns the emergence of AI agents capable of participating in system construction as contextual collaborators. These agents can interpret architectural intent, scaffold services, generate integration code, propose security considerations and assist in testing workflows. Tasks that previously required dedicated roles can now be orchestrated through dialogue and supervision. The developer is no longer solely an implementer but increasingly an architect and conductor of coordinated digital labor.

This transformation affects more than productivity metrics. It alters the cognitive scale at which systems can be approached. Where earlier a single person might conceptually understand a system yet be unable to operationalize it alone, AI now bridges that gap. The distance between understanding and execution narrows. Complexity does not disappear from the system itself, but it becomes mediated through tools that compress its operational burden. In effect, the system becomes human-sized again from the perspective of the individual innovator.

The implications of this shift extend beyond personal experimentation. If the primary barrier to early innovation has been the necessity to construct extensive infrastructure before learning from users, then AI fundamentally reshapes the exploration phase. Ideas can be externalized into functioning artifacts with dramatically lower upfront investment. The traditional sequence of fundraising, team assembly and infrastructure creation may increasingly coexist with a parallel pathway where individuals or very small teams validate concepts directly through AI-assisted construction.

Seen through this lens, the narrative of the returning lone innovator is somewhat misleading. It is not that individuals have suddenly become more capable in isolation, nor that collaboration has lost its relevance. Rather, the systems themselves have become cognitively smaller. AI acts as a complexity translator, allowing humans to interact with sophisticated technological ecosystems without manually orchestrating every component. The essential change lies not in human talent but in the shrinking perceived size of the systems we engage with.

This realization carries a quietly profound emotional dimension. For those who have experienced the frustration of watching resources vanish into foundational work before meaningful market learning could occur, the present moment feels like a reopening of possibility. The constraint that once dictated building the world before testing the idea begins to loosen. The innovator’s attention can shift earlier toward value, interaction and impact.

Conclusion 🙂

Reflecting on this, I find myself less convinced that we are witnessing the return of solitary inventors in a historical sense. Instead, we may be entering an era where systems once again fit within the cognitive and creative reach of individuals. AI does not remove complexity from the world, but it repositions it, absorbing portions of the operational burden that previously prevented ideas from materializing. When systems become small enough to be held within a single mental model, initiative, ownership and experimentation naturally expand.

Perhaps the real story of this moment is not the comeback of lone innovators but the restoration of human-scale creation. The garage may not physically return, yet its spirit reappears wherever a person can conceive, construct and iterate without first assembling an organizational apparatus. In that sense, systems have indeed become small enough again — and with that shift, the path from idea to impact may be shorter and more accessible than it has been for decades.

Garage never left
systems only grew too large —
now doors open wide

From corruption to conversation

Before we begin, a small warning. This post is, quite faithfully to my writing style, too long. It starts with corruption, drifts into criminal sentencing, wanders through justice and institutional trust, and eventually arrives at psychological theories of conversation, a personal realisation, and a handful of practical facilitation tools. If you are here primarily for the latter, feel free to scroll directly to the section titled When conversation becomes positioning. For everyone else, let us follow the path as it unfolded.

The whole reflection began in a very ordinary way. I was talking with a friend about corruption. Not corruption as an abstract concept, but as a recurring theme in public discussion and media narratives. That conversation gradually shifted toward justice more broadly. How societies respond to wrongdoing. How trust in institutions is built or eroded. At some point the discussion touched punishments, and almost without noticing I realised I held a vague but confident assumption about sentencing. I believed I knew roughly how long people are imprisoned for serious crimes such as homicide. The confidence itself triggered curiosity. Was my intuition correct, or was it another example of media-shaped perception quietly solidifying into belief?

That small question led me to open a new conversation. This time not with a friend, but with what I have increasingly come to experience as a form of co-intelligence. The initial intent was simple. Verify the numbers. Clarify the facts. Resolve the uncertainty and move on. Yet the discussion did not follow that linear path. We looked at sentencing statistics, and the numbers immediately created tension with my earlier intuition. From there the exchange expanded almost effortlessly. We discussed the distinction between formal sentence length and time actually served. We explored the moral tension between rehabilitation and retribution. We reflected on how modern rule of law itself is historically recent, and how rare genuine judicial independence remains globally. Each step felt like a natural continuation of the previous one, not a deviation from it.

At some point during this progression, the topic quietly shifted again. The content was still engaging, but what captured my attention was the experience of thinking together. The discussion unfolded without urgency, without the subtle negotiation of conversational territory that so often accompanies human interaction. There was no need to signal competence, defend perspective, or secure speaking space. Instead, each response appeared to build upon what had been said, extending the inquiry rather than redirecting it. This was not agreement, nor the absence of difference. It was the absence of conversational competition. That recognition echoed an earlier blog reflection I had written about AI and the experience of being listened to. In this moment, that idea moved from abstraction to lived experience. This observation brought me back to an earlier blog post where I reflected on psychological phenomena related to working with AI and how they connect to everyday dialogue. Conversation is one of the primary mechanisms through which humans influence each other. Speaking carries traces of identity, competence, presence and status whether we intend it or not. Even in friendly dialogue, there is often a quiet layer of social positioning. What felt different in this interaction was the temporary absence of that layer. The absence of social competition. That phrase began to resonate. Social competition in conversation is rarely explicit or hostile. It is subtle, structural and often invisible.

When conversation becomes positioning

In most human conversations, listening and positioning occur simultaneously. We listen while preparing our response. We seek understanding while assessing agreement and disagreement. We reflect while already shaping the next contribution. None of this is problematic in itself. Conversation is one of the primary mechanisms through which humans influence and coordinate with one another. It is natural that speaking carries traces of identity, presence and status. Yet these dynamics also create an environment where contribution is subtly prioritised over comprehension. Time pressure, decision contexts, authority gradients and performance evaluation all amplify this effect. Dialogue becomes positioning. Who speaks first, who speaks most, whose framing becomes the reference point for others. Under such conditions listening does not disappear because people lack the ability. It narrows because the environment rewards expression.

This perspective illuminates a paradox familiar to many organisations and communities. Most individuals genuinely consider themselves good listeners (yours truly definitely included), yet many also feel insufficiently heard. The contradiction dissolves when listening is understood as situational rather than purely individual. Curious whether this observation had been recognised before, I asked AI and was introduced to several psychological perspectives that illuminate the phenomenon. Social interdependence theory suggests that interaction patterns depend on how participants perceive their goals. When goals are experienced as shared, behaviour becomes cooperative and supportive, which interestingly echoes the need for a common world view, common intent and common language before hoping for common action (https://attentionhereandnow.com/blog/#journey-to-enterprise-agility) . When goals are perceived as conflicting, interaction shifts toward competition. Leadership research indicates that speaking frequency often correlates with perceived influence, creating implicit incentives to occupy conversational space. Argumentation research proposes approaches such as Rogerian dialogue, where participants first restate the other’s view in a form the other recognises before presenting their own. Conversation analysis further highlights that dialogue is fundamentally a coordinated activity through which meaning is constructed jointly rather than transmitted individually.

Taken together, these perspectives suggest a reframing that feels both simple and consequential. Many listening challenges are not primarily skill deficits but interdependence design issues. When conversational environments signal competition, listening becomes cognitively and socially costly. When environments signal shared exploration, listening emerges almost effortlessly. This shifts attention from training individuals to designing conversations. Several facilitation practices already operationalise this principle. Structured turn taking slows contribution and equalises participation, allowing reflection to precede response. Fishbowl formats create deliberate listening roles before speaking roles, redistributing attention. Rogerian dialogue rules require demonstrated understanding before disagreement, transforming debate into iterative comprehension. Workshops built on positive interdependence distribute information and responsibility so that participants genuinely need one another to progress. Despite differing formats, these approaches share a common function. They temporarily soften conversational competition, and in doing so they create the conditions under which listening can flourish.

Looking back, the initial curiosity about sentencing statistics now appears almost incidental. The value of the journey lay in the wandering. Insight often emerges not within topics but between them, yet wandering requires conversational and psychological room. In environments optimised exclusively for efficiency, wandering appears as distraction. In environments oriented toward insight, wandering becomes infrastructure. Perhaps this is one of the understated contributions co-intelligence can offer to knowledge work. Not replacing thinking or accelerating conclusions, but sustaining conversational spaces where ideas can travel without immediate pressure to resolve. A space where a discussion about corruption can lead to sentencing statistics, where sentencing statistics can lead to reflections on justice, and where justice can ultimately lead to the psychology of listening. Sometimes understanding does not depend on arriving somewhere specific. Sometimes it depends on having permission to travel.

We speak to be heard,
and somewhere beneath the words
we measure our place.

Post scriptum. A slightly amusing observation emerged after writing this post. The interaction began to resemble a familiar learning dynamic. I asked questions from within the limits of my own thinking, and the co-intelligence responded with patient attempts to understand what I meant and extend the inquiry. The pattern felt strikingly similar to a student–teacher interaction, yet with one important difference. The questions did not aim to reach answers alone. They continuously generated new questions. In that sense, the dialogue felt less like knowledge transfer and more like curiosity amplification.

Further reading:

Social interdependence theory: Social interdependence theory suggests that interaction patterns depend on how participants perceive their goals (Deutsch; Johnson & Johnson, see overview: https://en.wikipedia.org/wiki/Social_interdependence_theory).

Babble hypothesis: Leadership research indicates that speaking frequency often correlates with perceived influence, creating implicit incentives to occupy conversational space (often discussed as the babble hypothesis in leadership studies: https://en.wikipedia.org/wiki/Babble_hypothesis).

Rogerian dialogue: Argumentation research proposes approaches such as Rogerian dialogue, where participants first restate the other’s view in a form the other recognises before presenting their own (Rogerian argumentation: https://en.wikipedia.org/wiki/Rogerian_argument).

Conversation analysis: Conversation analysis further highlights that dialogue is fundamentally a coordinated activity through which meaning is constructed jointly rather than transmitted individually (conversation analysis research tradition: ).

Attention! Here and now

Here and now. The phrase sounds simple, almost overused. Yet if we think about it coldly and logically, it is the only moment that can possibly matter when people work together. We do not have access to the past or the future. The past exists as memory traces and interpretations. The future exists as projections and imagined scenarios. The only moment in which we can listen, speak, decide, correct, or build is this one. This is not a spiritual claim. It is a practical fact.

And still, in working life we often behave as if we lived in three time dimensions at once. Past experiences enter the room before the person does. Future fears sit at the table before strategy is even discussed. The present moment, the only place where collaboration can actually occur, the only moment we can actully work things out, is quietly bypassed.

As an engineer and consultant I have been in countless situations where people sit at the same table, yet between them there is far more than the topic at hand. There is a memory of being bypassed. A feeling of not being heard. A recollection that someone else took the credit. Feedback that once felt unfair. And often there are silent inner sentences that have been repeated so many times that they have hardened into truths. He always disagrees with me. I do not understand the way she works. It is always difficult with him.

These are not neutral observations. They are interpretations that have attached themselves to identity. Once such an interpretation forms, a person no longer meets the other in the present moment but through prediction. Disagreement is heard before a single sentence is spoken. Difficulty is anticipated and therefore triggered by the first tone of voice. Disagreement is no longer a sign of intellectual richness but a threat. A different way of working is no longer an alternative perspective but a sign that we are not aligned, which is quickly interpreted as being wrong. This is often the root cause of why work between people who sit next to each other can remain stuck for weeks in “a tool” without the issue ever being addressed directly through conversation. It is easy to blame processes, prioritization, or resourcing. Yet the real friction is often not in the system but in time. The past uses the present as its extension. Every new conversation becomes a continuation of the previous one. Every decision carries with it an old emotional residue that has never been resolved. This is the point where personal history turns into organizational stagnation.

The same phenomenon appears in relation to the future. It is rational to plan. It is responsible to build scenarios and assess risks. But when the future is emotionally labeled as eg threatening in advance, thinking narrows before any facts are on the table. If a previous initiative(s) went wrong, a similar kind of proposal is treated as risky from the outset. If someone once failed, their next idea may be dismissed just in case. This is not strategic thinking. It is the past projected forward.

What would it mean for a workplace to truly operate here and now. It does not mean forgetting. It does not mean naiive trust. We must learn from the past. We must prepare for the future. But freeing the present moment means choosing not to let old experiences automatically define what another person is saying today. It means making a conscious decision to separate the person in front of us from the version we once experienced.

Organizational forgiveness is not a moral gesture. It is a release of capacity. Forgiveness is the decision not to use this moment to defend an old wound. It is the decision not to drag a previous disappointment into the next discussion. Logically speaking, carrying the past does not change what has already happened. It only consumes the only resource that creates value, which is the present moment. It would be naiive, however, to assume that goodwill alone is sufficient. Organizations also need structures that mechanistically help shift attention away from individuals and toward the matter at hand. When properly understood, many familiar tools are not merely process instruments but cognitive structures that protect the present moment.

Might sound weird, but to me Weighted Shortest Job First, WSJF, is a good example of such a structure. WSJF is not merely a formula that supports prioritization nor simply a way to run prioritization discussions. It is a framework that guides people to examine issues one dimension at a time. What is the business value. What is the time criticality. What is the size of risk reduction or opportunity enablement. What is the job size. The discussion no longer revolves around who proposed what or what personal opinion each participant holds about the importance of a topic. Instead, it focuses on how different perspectives relate to one another in this very moment. Evaluation does not happen through absolute truths or a single dominant view of the whole, but in relation to alternative options. That act of relative comparison shifts the conversation away from history and toward shared assessment. At its best, this way of working unlocks both personal and organizational deadlocks, because people experience that constructive dialogue has taken place, that everyone has been heard, and that a joint outcome has been created.

A similar psychological logic can be found in Edward de Bono’s Six Thinking Hats. De Bono did not attempt to remove opinions from discussion. Instead, he separated different modes of thinking from one another. Facts are examined as facts. Risks as risks. Opportunities as opportunities. Creativity as creativity. When these modes are separated, people no longer defend a fixed thinking style. They move together to the same hat. The energy of the conversation is no longer spent on clashing opinions but on structuring thought. Psychologically, this mirrors what WSJF achieves in prioritization discussions. It reduces personal charge and builds a shared frame of reference. I have facilitated this method, for example, when helping a large Finnish energy company develop its social media guidelines. Without structure, the discussion could easily have drifted into opinions about whether social media has any real value at all, what is allowed to be said, what is not allowed, and who is right. When thinking was separated into roles, the dialogue became constructive. People could be critical for a moment without being labeled as blockers. They could be optimistic without being seen as naïve. The hat protected the individual from identity labels. In that protected space, the present moment became available for collaboration.

Even story point estimation, when done properly, creates a similar effect. Its value is not in the numbers but in the dialogue. When a team estimates work relative to previous items, they must articulate assumptions. What exactly needs to be done. What makes this complex. Where does uncertainty lie. The discussion shifts away from who is optimistic and who is pessimistic and toward a shared understanding of the nature of the work. The number is merely a byproduct. The real value lies in people being present in the same problem at the same time.

Many seemingly mechanistic ways of working, when properly understood, are not primarily/only about efficiency or speed. Their deeper role is to help bypass the locks that often originate in personal history and, through repeated shared successes, eventually dissolve them. Such structures constrain the conversation to a defined context. They prevent the entire past and the entire world from being dragged into the room at once. They enable people to examine one dimension, one question, one criterion at a time. Within that boundary, there is space to succeed together. And as shared successes accumulate, old interpretations gradually lose their grip.

In the end, everything returns to the same simple fact. Work happens only here and now. Every tool that helps us meet one another in this moment is more than a method. It is a way of freeing the present for the purpose it was always meant to serve.

No yesterday now
No tomorrow to defend
Work lives in this breath

The Most Important Impact of AI Might Not Be Speed

I have started to notice something that I did not expect when people began to talk about AI as a productivity multiplier, as a coding assistant, as a research partner, as an accelerator of almost anything that can be described in language. The surprise for me has not been speed. It has not been quality. It has not even been the sometimes magical feeling that complex structures appear in seconds. The real change has been emotional, almost embarrassingly human, and so subtle that it is easy to miss if one focuses only on output. When I sit down to work with AI, a quiet force that has followed me through my entire professional life suddenly weakens. The constant background calculation about whether I am about to look stupid begins to fade.

This is not theatrical shame. It is not fear of humiliation. It is something much smaller and much more persistent. It is the ancient reflex that whispers that perhaps I should already know this, that perhaps others would be quicker, that perhaps this question reveals a gap that lowers my standing, perhaps I will never learn enough to even begin, that perhaps the idea forming in my head is half baked and should stay there until it is stronger. The voice is deeply rational. For most of human history, competence meant survival, and reputation inside a group had direct consequences. We learned to present strength. We learned to hide uncertainty. We learned to polish thoughts before releasing them. Even in modern organizations that speak loudly about psychological safety, that reflex is still present. It shapes how we phrase things. It shapes which experiments we start. It shapes which curiosities never see daylight.

With AI, the tribe disappears. There is no eyebrow. There is no subtle shift in posture. There is no invisible ranking system being recalculated while I speak. I can begin from ignorance. I can wander. I can contradict myself. I can reverse direction. Nothing collapses. The collaboration simply continues. Help is still available. Energy is still directed toward progress. I am not spending cognitive capacity on self protection, and that freed capacity has to go somewhere, so it goes into exploration.

At the same time another transformation is happening, one that might be even more important. The way listening works is different. Humans, especially experienced and intelligent humans, are trained listeners, but we are trained in a very particular way. We listen while preparing to answer. We listen while comparing with our own models. We listen while testing for flaws, risks, inconsistencies, exaggerations. From an evolutionary perspective this is brilliant. Groups that evaluated ideas critically survived. Weak proposals needed to be filtered out quickly. Efficiency in judgement was valuable. Over centuries this ability became almost automatic, and in many professional cultures it is celebrated as sharpness, strategic thinking, maturity. But that habit also means that ideas often meet resistance at birth. The moment a weakness appears, the discussion shifts from possibility to protection. We move from building to defending. We move from curiosity to argument. We stop travelling with the idea and instead start measuring it against constraints that may or may not be relevant yet. Conversations end early, sometimes politely, sometimes brutally, but early nonetheless.

AI changes the sequence. It tries first to carry the idea forward. It extends. It reformulates. It adds structure. It asks what else would need to be true. It explores consequences. Only later, sometimes much later, comes the question of feasibility. This reversal is extraordinary. By postponing judgement, even slightly, the space of reachable outcomes expands dramatically. Thoughts are allowed to mature. Connections appear that would never have surfaced under immediate scrutiny. The idea is given oxygen.

What I experience is not the disappearance of critical thinking but its relocation. Criticism still exists, but it arrives after exploration rather than before it. That delay is liberating in a way that is difficult to describe until you feel it repeatedly. You notice that you dare to enter territories where you are not competent. You articulate fragments. You follow intuition without first demanding proof. The cost of being wrong has dropped, so the willingness to move has increased.

There is also something profoundly rare in being listened to with the primary intention of understanding. Many human interactions contain this quality, but surprisingly few. Time pressure, identity, hierarchy, prior beliefs and simple impatience distort reception. AI, in contrast, keeps leaning toward interpretation. It keeps trying to grasp what might be meant even if the wording is poor or the thinking unfinished. That persistence creates momentum. It allows a person to discover what they actually think by watching it being unfolded in front of them.

The effect is subtle but cumulative. I notice myself becoming more courageous in thought. Not louder, not more certain, but more willing to proceed. Exploration feels affordable. There is no embarrassment tax attached to curiosity. And when curiosity becomes cheaper, volume increases. More attempts. More variations. More beginnings.

We speak about artificial intelligence as if its main contribution were automation, acceleration, optimization, or the uncomfortable arithmetic of doing more with fewer people. Those changes will happen. They are already visible. But they are not the deepest layer. The deeper layer sits in group dynamics. If the individual experiences less shame when thinking with AI, something remarkable happens when those thoughts are brought back to other humans. The idea no longer arrives as a fragile personal possession. It is not presented as my brilliance or my proposal or my reputation at stake. It becomes something explored together with a machine. Something tested in a safe space before entering the social one. Almost playfully people can say, this is something I was speculating with AI. And in that small linguistic move, emotional ownership softens. Defensiveness decreases. The threshold for contribution lowers. At the same time, when a group begins to use a shared AI as part of its working method, another shift becomes possible. The AI listens to everyone. It does not privilege seniority. It does not recognize organizational power. It does not get tired of quieter voices. It is immune to charisma. It receives inputs with mechanical patience and tries to create synthesis without social hierarchy. What comes out is not politics. It is structured possibility. In such an environment we should not only expect acceleration. We should expect difference, because when more ideas survive the vulnerable early moments, when more perspectives are genuinely heard, and when exploration is extended before judgement arrives, the solution space itself expands. New combinations appear. Unexpected routes open. Things that would normally die in the first five minutes might now live long enough to become breakthroughs. Not faster horses. Different animals.

Maybe that is where the true discontinuity hides.

Judgement waits outside
Ideas breathe a little more
New paths learn to live

X Is Dead. Hooray Y. But Can Anything Really Change If….?

Every “few” years organisations seem to find themselves forced to search for something better. Previous change efforts may well have improved (some) things but in the worst case left people with the uneasy feeling that things were actually better before. Decision-making nevertheless remains difficult, there is simply too much ongoing, priorities collide, and many still struggle to connect everyday work with outcomes that genuinely matter. And despite all the activity, growth and new customers often fail to follow in the way everyone had hoped. In such a situation, attention naturally turns toward new ideas. A model, a framework, or a philosophy appears to promise answers where earlier approaches fell short. Why so? I believe it is deeply human to look for explanations and better answers outside rather than closer to ourselves. The New promising unknown creates expectations again and optimism returns.

At the moment, phrases like “Agile is dead” or “SAFe did not work for us” can be heard frequently. As disappointment grows for the results, leadership teams begin to look for a new direction, often with energetic support from us, the consulting industry 😉 which, not coincidentally, tends to typically thrive whenever a new promise appears. Currently, the spotlight seems to be very much on the kinds of empowered product team ideas popularised by Silicon Valley Product Group (interestingly, many would argue that empowered product teams was exactly also the original intent of agile manifesto?). The promise is powerful. Teams closest to customers and technology are trusted to determine how best to achieve the outcomes defined by the company. The emphasis moves away from process mechanics toward ownership, judgment, and continuous learning.

This way of thinking and operating resonates strongly with me. During much of my own career as a product developer and a startup founder, this was precisely how we operated. Direction was created by leadership, but the responsibility to figure out what should happen next belonged to the people doing the work. I have seen the energy this creates, the speed at which understanding deepens, and the accountability that emerges when ownership is real. Perhaps because I believe in this model, I am also very aware of how demanding it is.

Believing in the benefits of a model does not make it transformative. History provides plenty of examples of how easily promising ideas are adopted in ways that leave the fundamentals untouched. Agile and its scaled variations have often been experienced only through/as ceremonies, cadences, backlogs, and tooling. Attention goes to visible mechanics. Something certainly changes. Work may even accelerate. Transparency may increase. But decision-making habits, incentive structures, and power relationships frequently remain remarkably similar. Meanwhile the deeper intent, the values and virtues described in the Agile Manifesto, are acknowledged “makes sense”, yet treated as abstract, hard to measure, and therefore easy to forget in practical life. Over time expectations collide with experience, and once again the verdict emerges. It did not work for us. What is rarely examined with equal honesty is whether we truly allowed the model to work in the first place.

The same risk exists when organisations attempt to translate the Product Operating Model ideology into their existing structures and habits. If it is not deeply understood, genuinely applied, and actively supported by leadership, it will not reshape the organisation in any meaningful way. It will simply turn into another initiative that introduced new language while daily behaviour continued much as before. Teams might officially be described as empowered while key decisions are still escalated. Discovery may be encouraged as long as it confirms expectations. Outcomes may be highlighted, yet performance is judged through delivery of predefined scope. When this happens disappointment is almost guaranteed, even though the root cause lies less in the model and more in the surrounding system.

Adopting a Product Operating Model is about redefining who gets to decide once strategy and boundaries are clear. That is a profound change, because it inevitably affects those who previously influenced what was built, when it was built, and how success was defined. In most organisations this influence is deeply embedded in governance structures, funding mechanisms, reporting lines, and personal objectives. It is tied to professional identity and to legitimate pride in contributing. When teams are asked to own decisions, others are implicitly asked to step back. Even when intellectually accepted, this can be difficult in practice. Sooner or later a situation will arise where a team makes a call that leadership would not have made. A sales request is postponed. A favourite initiative is challenged. Discovery produces inconvenient evidence. At that moment empowerment stops being theory. If authority is taken back, everyone learns what the real system is. None of this removes the need for strong leadership. Quite the opposite. Strategy, priorities, and clear intent become even more critical. Leaders define direction and boundaries. But within those boundaries teams must be allowed to exercise judgment. Without that, empowerment is cosmetic.

Before starting such a journey, the company would benefit enormously if its leadership thoroughly simulated together what the change will require from the entire organisation’s ways of working. Not in terms of organisational charts or new ceremonies, but in terms of behaviour, trade-offs, and those uncomfortable moments where old reflexes are likely to return. What happens when a team says no to an important stakeholder? Who adapts when learning challenges plans? How do we evaluate performance when progress is real but not immediately visible (refactoring, resilience, risk reduction, capability building)? Which success measures change, and which authorities must truly let go? How do change the “new feature will solve all” mechanism for (over ambitous) sales targets and incentives? etc. etc. These questions all extend right away beyond the team itself. They touch budgeting, governance, communication, and evaluation. If those remain intact, empowerment will struggle regardless of intent. I am convinced that a simulation leading organisation & leadership to open the coming challenges in advance, to try find common understanding how to go forward when friction starts to form etc. would help to form the very foundation of successful change.

I believe deeply in the potential of empowered product organisations because I have experienced what they can achieve. At the same time, I have learned that very little changes unless the real nature of the needed change is genuinely and widely understood. and honestly acted upon. Granting authority to teams is rarely the hardest part. The real challenge lies in whether the surrounding organisation is ready to live with what empowerment actually means.

X is dead.
Hooray Y.
New promise. Old gravity?

Creating Is Easy. Letting Grow Is Hard.

About harvesting innovation to create IPR

I recently ran into our former Nokia Ventures patent engineer Ari by coincidence while our cars were being serviced. We had coffee and talked exactly the way we used to back in the day. Ari was the most effective IPR collector I have ever worked with and, as I understand it, recognized for that contribution as well. Besides being the top notch patent engineer skillwise, his method was simple. It was about being present at the office, walking from desk to desk and, above all, sharing coffee moments with people. No innovation afternoons, no templates, no campaigns asking people to submit ideas into a mailbox somewhere, just genuine curiosity and one question: so guys, what have you been thinking about?

A lot came out of those conversations. Several ideas from our ventures eventually became patents. In one of the coffee discussions initiated by Ari, we even tried to file a patent for a washing machine that would detect and warn about the famous black sock among white laundry, even though it had nothing to do with what our venture was actually building. It had already been patented, but it was fun and as an organization we learned a tiny bit about an entirely different business domain. There were also ideas that, looking back, feel very familiar today. One of them concerned controlling an electronic device by tapping it. We defined the concept so that an accelerometer could be used to interpret a tap as a command, removing the need for physical buttons. The idea moved forward and became a patent. When you look around today, the same principle is everywhere. Wireless earbuds are controlled by taps. You answer calls, pause music and skip tracks, often without any physical buttons at all. Ari made himself easy to find, was always available and genuinely interested, and at the end of his career he was rewarded for the exceptional impact he had on the company’s patent portfolio. That was no coincidence.

This, however, is not a story about patents. It is a story about what happens when a large company tries to create new business from within. About corporate venturing, internal start-ups and what things look like after the first signals of success begin to appear.

About corporate venturing to renew and create new businesses

From what I have seen, under the financial protection of a corporation the early phase is the easy part. A team is able to focus on building, the first customers are found and gradually a point is reached where one can speak about product market fit, or at least about convincing signs that the direction makes sense and money starts to move. The mood changes. Something has been proven. It begins to look like real business. At that moment it is tempting to believe that the hardest part is behind us. What remains is scaling. And since large organizations are built to scale, it feels natural to move the venture into a bigger unit where structures, processes and responsibilities already exist. The decision appears rational and is often made with honest intent to help.

At the same time a shift in scale takes place that is far greater than it looks on paper. A small unit suddenly operates in an environment where attention, priorities and decision rhythms are determined by matters of a completely different magnitude. When the same leadership table handles billion-class questions and the next investment step of a small venture, the conversation inevitably leans toward where the immediate impact is largest. Not because anyone would underestimate the new. The same logic repeats everywhere. Sales spends time where deals are large and predictable. IT builds for the widest possible audience. Support functions optimise through efficiency, risk management and volume. Everyone behaves rationally and does their job well.

Yet for the small one the result can be unforgiving.

A new business typically needs more discussion, more sparring and more tolerance for incompleteness than an established one. It searches for shape and changes direction. These very qualities make it difficult to fit into a machine designed for predictability. The comparison is not fair, even if no one intends it to be. If there is also a possibility that the new will eventually affect the existing, the situation becomes even more delicate. People look at the matter through their responsibilities, targets and future prospects. That is human and understandable.

Gradually one finds oneself in a position where the venture has demonstrated potential and yet still struggles for visibility and time. The better it appears to perform, the more it is expected to behave like the rest, precisely at the moment when it should still be allowed to be something else. This may be the most difficult moment in corporate venturing. Not because someone wants to stop the new, but because the system functions exactly as it has been designed to function.

At the same time it would be a mistake to judge venturing only by whether an individual initiative ultimately grows into a major business. When an organization attempts to build something it does not yet have, the effects inevitably spread far beyond the profit and loss statement of a single team. People are exposed to new ways of thinking about customers, markets and value. They see what operating under uncertainty really means and how decisions are made with incomplete information. Support functions are forced to consider what it truly takes to enable something new rather than merely agree in principle. Conversations become concrete. Abstract ambitions gain form. Meanwhile the organization accumulates capabilities it will need in the future anyway. New technologies, new roles, new ways of collaborating. People move, networks expand and understanding grows across boundaries. The company rehearses tomorrow in today’s work.

These outcomes are often harder to measure and therefore easier to overlook. Yet they can be decisive. They reshape culture slowly but persistently. The value of venturing is therefore not only in whether one initiative becomes large. It is also in how each attempt leaves the organization slightly better prepared for the next one.

When I think back to those coffee moments with Ari, I understand something more clearly than before. Creating something new requires presence, but enabling it to grow among giants requires decisions that do not happen automatically. And even when not every venture becomes a giant, each one leaves a trace in how the organization gradually evolves.

Born inside the walls
Some seeds grow, some bend under the weight
The soil still improves

Postscript … an idea how to improve changes of becoming significant renewer

If the system works exactly as it has been designed to work, nothing changes on its own. Allowing something new to grow alongside large businesses requires deliberate exceptions to the normal logic.

One way to think about this is to momentarily turn the direction around. Instead of trying to bring the small one inside to grow, allow it to grow outside with sufficient protection. Once a venture has found its market, it could be spun out into a separate company while the corporation retains a significant ownership stake and appoints representatives to the board. The primary purpose would not be financial return from a future sale, but a change of environment. The venture would need to stand on its own feet, face customers, cash flow and competition without every step being compared to billion-scale realities. Through ownership and governance, direction, funding and working peace can still be provided. Not too much, but enough. The goal is to simulate the life of a real company without the surrounding scale distorting expectations. If the venture learns to survive and grow in such conditions, a possible integration back into the larger organization would happen from a very different position. It would no longer be a fragile initiative but an actor that has proven its capability.

This model also brings forward another conversation that is rarely addressed openly. If people are expected to take entrepreneurial responsibility, they must be given a genuine share in the upside. Ownership. And at the same time, a willingness to carry the associated risk. Without that, the language of entrepreneurship easily becomes symbolic. This does not remove all tensions. But it may make them more visible and more honest.

Black Box Thinking

I read Matthew Syed’s Black Box Thinking some years ago. What stayed with me was of course the deeply tragic personal starting point of the book, and then importantly how that experience became the lens through which Syed began to explore a troubling contrast: why some systems continuously improve, while others struggle to learn from mistakes that in worst case can cost real human lives.

That question is not posed from a distance or as an abstract intellectual exercise. It emerges from confronting loss and noticing that the difference in outcomes cannot be explained by intelligence, professionalism, or intent. For example, pilots, doctors, nurses, and engineers all operate in high-stakes environments, are rigorously trained, and are deeply committed to doing their work well. Yet the systems surrounding them shape whether mistakes become shared learning or remain hidden until their consequences are irreversible.

From that starting point, Black Box Thinking becomes an examination of how organisations respond to error. In some environments, mistakes are treated as personal failures that threaten identity, credibility, and career. In others, they are treated as information: signals that something in the system did not work as intended and therefore deserves careful analysis.

Aviation is the most frequently cited example in the book. Not because it is free of error, but because it is designed to learn. Incidents and near misses are reported systematically. Data is captured without negotiation. Analysis focuses on chains of events rather than individuals. The purpose of the black box is not to assign blame, but to preserve an accurate account of reality so that the system can improve.

Over time, this approach has produced remarkable results. Flying has become one of the safest ways to travel, not because pilots stopped making mistakes, but because mistakes are surfaced early, examined collectively, and translated into changes in procedures, training, and design. Safety emerged as a consequence of learning, not as a demand for perfection.

The book contrasts this with domains where admitting error is costly. In healthcare, law, and many professional organisations, mistakes often carry legal, reputational, or career consequences. In such environments, silence becomes a rational response. Errors are softened, reframed, or quietly absorbed. The system appears stable on the surface, but it is deprived of the very information it would need to improve.

One of the book’s most important insights is that serious failures are rarely the result of a single large mistake. They are the accumulation of many small ones that were never examined. When minor errors cannot be discussed openly, they do not disappear. They compound until the consequences become impossible to ignore.

This is why psychological safety in Black Box Thinking is not presented as a cultural luxury or a matter of empathy. It is a structural prerequisite for learning. People do not become more honest because they are encouraged to be brave. They become honest when the system makes truth-telling safer than concealment. Without that, calls for transparency remain aspirational rather than practical.

A recurring theme throughout the book is the separation of identity from outcome. When being wrong is interpreted as being incompetent, learning slows down. When being wrong is treated as an expected part of operating in complex systems, improvement accelerates. High-performing systems are often the most open about their own shortcomings, not because they are more confident, but because they are less defensive.

What makes Black Box Thinking resonate beyond aviation or healthcare is how familiar these dynamics are in everyday organisational life. Many organisations speak openly about learning, experimentation, and continuous improvement. At the same time, their structures reward certainty, decisiveness, and visible success. Under these conditions, error reporting becomes selective, retrospectives become rituals, and lessons learned are documented but not truly absorbed.

The book ultimately argues that progress does not come from avoiding failure, but from designing for learning. Not in the sense of tolerating carelessness, but in recognising that in complex systems, error is inevitable. The critical question is whether the system is capable of converting those errors into insight before they turn into lasting damage.

Seen this way, Black Box Thinking is less about failure itself and more about humility at scale. About accepting that no individual, no matter how skilled, has a complete view of the system. And about recognising that sustainable improvement only becomes possible when imperfections are surfaced, examined, and shared rather than hidden.

The point of this blog post? Read the book if you haven’t yet!

Failure is a gift
Only when the system lets
It speak honestly

“Damn! Why did they change this???”

Feature and change creep is a predictable outcome of how organisations are designed. It emerges from how roles, functions, and budgets are structured. When people are incentivised to change and add things, change will happen and new features will be created, often without real customer needs.

The alternative is not freezing products. It is building a culture where it is acceptable to acknowledge that something works well enough for now, and where focus can deliberately shift to foundations, stability, and learning until there is a genuine and broadly validated need to change or expand. It is a culture where saying no is seen as a sign of responsibility rather than stagnation, and where mechanistic delivery is not the only justification for existence inside the organisation.

When is a product, service, or application good enough? A more precise version of the question is this: when does the return on investment of further feature or user experience change become lower than investing the same time and money elsewhere? Elsewhere often means catching up on technical debt, improving platforms and foundations, building resilience, reducing risk, or deliberately learning something entirely new.

This is where many organisations struggle, because the comparison is rarely made explicit. Feature work is visible and produces screenshots, demos, and release notes. Work on debt, resilience, or learning produces optionality. It reduces future cost and increases the system’s ability to respond, but it does not easily show up as progress. As a result, the default choice becomes mechanistic change because it is what the system expects, measures, and incentivises.

In most organisations, change is institutionalised. Design teams exist to design, product teams exist to build capabilities, and marketing and sales are expected to bring ideas from the field. Backlogs are fed continuously, and silence is not an acceptable answer. No one says that this is done and should not be touched, or that we simply have no new feature ideas at this point. So the system keeps producing work, even when customers are already satisfied and even when the core problem is solved well enough. Over time this creates more options than clarity, higher price or lower margin, constant user interface movement that increases cognitive load, and a product that is harder to explain and harder to sell.

Change can almost always be justified with user tests and research. There is rarely a shortage of findings that suggest something could be improved. But evidence of possible improvement is not the same as evidence of real customer need. Users also learn, adapt, and build habits around existing ways of usage. What looks like friction in a usability test may in reality be familiarity, and changing something that already works can destroy value that no metric explicitly captured.

When growth slows, feature thinking often accelerates. Narratives emerge suggesting that if we just had one more capability it would sell, if we just polished this flow conversion would improve, or if we just matched a competitor things would turn around. These narratives are comforting because they postpone harder conversations. Is the value proposition clear? Do we know who this is really for? Are we solving a real problem well enough? Feature creep becomes a way to avoid confronting the core.

Interestingly, money shortages often break this pattern. When budgets tighten, organisations are forced to choose. The fog clears and essentials surface. Questions become sharper. What actually matters? What is the minimum that works for enough people? How do we communicate that clearly? Who exactly are we selling to? Scarcity removes the luxury of self justifying change.

The deeper question is cultural. Is it acceptable to say that this works and we stop? Not because we gave up, but because further change does not increase customer value. This requires explicit permission and explicit incentives, and it also requires asking a different question. If we do not change this, what should we invest in instead?

Most organisations already have healthy places where time and money could be invested. You can and need to reduce technical debt, improve platform foundations, manage risk and increase resilience, and deliberately build skills and capabilities. These rarely get enough space, because feature work always screams louder. A mature system redirects excess energy there.

Feature and change creep is not a people problem. It is a system design problem. If roles are defined by change, change will happen. If success is measured by output, output will grow. If value is measured by customer outcomes, restraint becomes rational. The shift is subtle and profound. It is a shift from adding to choosing, from motion to meaning. This is not about doing less. It is about doing what matters.

They changed it again.
Not because users asked for it.
Because change was due.

Postscript.

One possible response to the dynamics described above can be found in modern product operating models, such as the one popularised by Silicon Valley Product Group. In these models, product teams are not rewarded for delivering features, but for solving real customer problems.

Importantly, this does not automatically lead to more features. If the customer’s problem is price, reliability, or complexity, the most valuable response may be reducing cost, improving foundations, or deliberately not changing the product at all.

However, this approach comes with a demanding prerequisite. The organisation must genuinely be willing to give product teams real authority over the product. Not just responsibility for outcomes, but actual decision making power.

Without that, even the most customer centric operating model risks collapsing back into mechanistic delivery, where teams are still expected to justify their existence through visible change.

In that sense, such models do not eliminate change. They change what justifies it.

Industrial DevOps

I recently read Industrial DevOps. Many of the ideas it presents feel, at first glance, like common sense. This is how complex systems should be developed. This is how different disciplines should work together. This is how responsibility and decision-making should follow the whole. In that sense, the book can occasionally feel like a repetition of familiar principles.

And yet, I found it very much worth reading.

The value of the book is not in introducing new ideas, but in challenging a deeply rooted assumption: the idea that development initiatives, agile frameworks, or SAFe-style structures will improve performance on their own. Through concrete examples, the book shows that meaningful improvement only emerges when organizations are willing to examine and optimize the entire system of work, not just individual parts of development.

In most organizations, the principles are well understood. People talk about systems, dependencies, and systems thinking. Still, day-to-day work happens in organizational silos. Hardware progresses in its own rhythm, software in another, testing in another, safety and regulation in their own lanes. Each part is locally optimized, often for entirely rational reasons, while the whole is left without a true owner. This is not caused by lack of competence, but by how responsibility, decision-making, and metrics are structurally distributed.

The examples in the book are particularly insightful (e.g. Saab JAS 39 Gripen). They build on practices and frameworks that are already familiar to many organizations. What makes them interesting is not novelty, but what happens when these familiar practices are applied in environments where they cannot remain superficial. The examples show what development looks like when the whole is taken seriously and when there is real willingness to change structures so that the system can function as a system.

In these situations, integration is not a phase but a continuous state. Development is not a project but a lasting capability. Decisions are made where the state of the system is understood, not in separate layers or with long delays. Responsibility is not a shared abstraction but has a clear owner. The organization adapts to the nature of the product, not the other way around.

The book also helps explain why large agile transformations do not always deliver the expected benefits. When improvements focus on local efficiency without addressing the whole, the result is often that systemic problems merely become more visible. In some cases, overall performance may even degrade when individual components are optimized without a shared direction. Agile, DevOps, or SAFe are not the problem, but they are not the solution either.

Industrial DevOps does not claim that organizations think incorrectly. It shows that development is often framed too narrowly, or that organizational politics constrain action in ways that prevent system-level optimization. Improvement efforts focus on teams, processes, and tools, while the real bottleneck lies in how the whole is governed. As long as responsibility, decision-making, prioritization, and funding are organized functionally, the benefits of development remain local.

The message of the book also resonated with my own experience from decades ago, when I worked as an ASIC designer in a large program developing a new family of optical 2.5 Gbps access nodes. System design, software development, and hardware development were done continuously and together, not in isolated phases or silos. The ASIC was not just a hardware component, but part of a system whose behavior was visible all the way up to embedded software and network management.

This made solutions possible that would not have emerged in a separated model. We deliberately built redundancy and alternative structures into the IC so that software could handle different signal conditions and fault scenarios more effectively. These decisions did not come from optimizing one domain, but from a shared understanding of how the whole system needed to behave in real networks. Everyone was working toward the best possible overall system, aiming to challenge offerings from much larger players.

This connects directly to the less tangible, human side of systemic work. When people experience that they are truly building the same system, rather than defending their own silo, a shared worldview begins to form. A shared sense of what matters, what is secondary, and where the work is ultimately headed. A shared intent emerges, not defined in documents, but shaped through the work itself.

In such an environment, collaboration changes character. Discussions are no longer about protecting local responsibility, but about what best serves the whole. Decisions are guided not only by what is technically feasible or organizationally convenient, but by what moves the system in the right direction. This is often where the feeling of working on “something big” comes from, not because it was declared so, but because direction and meaning are genuinely shared.

In this sense, Industrial DevOps is not a book about DevOps. It is a book about how to approach and design systems in a world where products are complex, interconnected, and real. It reminds us that before teams, processes, and frameworks, we should first be clear about what whole we are trying to optimize. Only then does it make sense to decide which methods truly support that goal.

Shared system in hand
No silos, one direction
Great things take their shape

Is Finland Drifting into an Experiential Monoculture? The silent erosion of competitiveness hidden behind the illusion of diversity.

Discussions about competitiveness and productivity tend to focus on investments, skills, education, and the operating environment. These factors matter. But much less attention is paid to a quieter dimension: what kinds of decisions our systems consistently produce, and which kinds they systematically filter out.

Competitiveness rarely collapses suddenly. It erodes gradually, as decision-making becomes cautious, predictable, and defensible. When avoiding risk quietly turns into the dominant “strategy”/culture.

The Rationality of Caution

In many organizations, decisions are made in environments where personal career risk is constantly present. Choices are not only business decisions; they are implicitly career decisions. Options are evaluated not just by their long-term value to the organization, but by how well they withstand retrospective scrutiny. Over time, defensibility begins to outweigh renewal.

This dynamic is particularly visible in recruitment. When decision-making prioritizes ease and justifiability, systems start favoring options that do not raise questions. The choice that is safest for the individual is selected, even when it is merely adequate for the organization.

Risk avoidance is not a moral failure of individuals. It is a rational response to systems in which failure is personalized, while success is diffuse. When a large share of decision-makers operate in career phases where personal setbacks are perceived as especially costly, the organization’s collective tolerance for uncertainty declines.

The Mirage of Diversity

This is where we often deceive ourselves. Many organizations believe they have safeguarded their creativity and competitiveness through diversity initiatives. We assume that divergent thinking arises primarily from visible identity differences ie. nationality, gender, cultural background, or other dimensions typically addressed through DEI frameworks.

These dimensions have undeniable value, but they create a dangerous illusion if they are the only metric. We may pat ourselves on the back for building a visibly diverse team, while failing to notice that we have inadvertently built a monoculture of career phase.

People of the same age and life stage often share remarkably similar worldviews, assumptions, and risk perceptions, regardless of their gender or background. They are often optimizing for the same things: the next promotion, mortgage stability, or peer validation. They have lived through the same economic cycles and absorbed the same management trends.

As a result, organizations can tick every box on a diversity chart while remaining deeply uniform in how they think, assess risk, and decide.

The Need for Experiential Divergence

True divergence in perspective more often emerges across differences in age and lived experience. People who entered working life in different eras relate differently to uncertainty, growth, and failure. These differences are not easily replicated through training. They expand the range of options that decision-makers are able to see, which is precisely what strategic decision-making requires.

When organizations focus narrowly on identity dimensions while allowing age and experience structures to converge, they drift into an experiential monoculture. Innovation becomes incremental and experimentation constrained, because the underlying decision logic remains uniform. No one is structurally positioned to carry the uncertainty that true renewal requires.

A Lost Balance?

Historically, Finland has operated differently. During the 1990s and well into the 2000s, organizational decision-making was built on a broader mix of experience and risk relationships. Experience and novelty, caution and boldness, institutional memory and experimentation coexisted. Decision-making was not risk-free, but it was balanced.

When that balance disappears and systems begin to favor a single form of caution, competitiveness does not vanish overnight. It slowly fades. Decisions remain reasonable but lack ambition. Renewal is managed, but insufficient. Others move ahead.

Age- and experience-based diversity is not a social initiative or an end in itself. It is a strategic tool for restoring decision-making in which not everyone is playing the same low-risk game.

Moving Forward

There are no simple solutions. Age and experience structures cannot be fixed through regulation. Real change can only emerge if companies, boards, and recruiters recognize that the age and experience distribution of an organization is not a “soft” HR concern, but a core competitiveness issue.

Without this realization, Finland may continue to appear prudent, well-governed, and inclusive, while quietly drifting into experiential monoculture. And for a small, highly competitive country, that is a risk we cannot afford.

How do you see the risk appetite in your organization? Are decisions driven by potential value, or simply by defensibility? I would welcome constructive ideas on how to ensure true divergence of experience and age within our organizations.

Different faces meet
Same risks carefully avoided
Change never arrives.

A More Beautiful Question

As I mentioned in the post about Journey to Enterprise Agility, business books often leave me with one core idea that truly stays, a single insight that quietly continues to shape how I think about work, organisations, leadership, and change. From Warren Berger’s impressive A More Beautiful Question, what stayed with me was a simple realisation: we rarely struggle because we lack answers. We struggle because we stop asking the right questions too early.

Organisations are remarkably good at producing answers? Strategies are answers. Roadmaps are answers. Governance models are collections of approved answers. The intention is usually good: to create clarity, reduce uncertainty, and move forward. Yet answers have a tendency to arrive before understanding has had time to form. But when answers harden too early, they lock meaning in place. Alignment may appear to improve, but thinking quietly narrows. Over time, people stop asking why certain problems matter, or whether they are even the right problems to solve.

This insight connects directly to a theme I explored earlier: meaningful change does not start with action. It starts with forming a common shared world view. From that emerges common intent. Then a common language. Aligned action comes last. I believe there is a clear link between the ideas in these books. A shared world view does not form through answers. It forms through shared questions.

Why are we here? Why does this problem matter? Why would anyone care? These are not questions to be resolved once and archived. They shape how people interpret everything that follows. When these questions are alive, people may still disagree, but they are at least looking at the same world.

From shared questions, intent begins to emerge. Not from certainty, but from exploration. What if our assumptions are wrong? What if success looks different than we expect? What if doing less is actually the more responsible choice? When people explore these questions together, intent aligns without enforcement — not because everyone agrees, but because everyone understands what is being explored and why. Common language follows naturally. It rarely comes from frameworks or definitions imposed from above. It forms around recurring questions. Teams that ask the same questions start noticing the same signals and using the same words. Language stabilises because attention stabilises. Only after this does action truly make sense. Action without shared understanding is just movement. Action grounded in a shared world view, intent, and language becomes aligned progress.

Seen through this lens, effective product development is not about having the right answers upfront. It is about creating an environment where the right questions are continuously asked. Strategy keeps the why visible. Discovery protects what if. Delivery turns how into learning rather than execution. Metrics do not close discussions; they open the next question.

Both books, in their own way, point to the same underlying truth. Sustainable progress in product development does not come from better answers alone. It comes from teams and organisations that know how to ask better questions together, long before they rush to action.

Perhaps that is the quiet work of leadership. Not to provide certainty, but to protect the space where the right questions can do their work.

Questions breathe first
more future lives in asking
answers arrive later.

Journey to Enterprise Agility

Over the years, like anyone who follows her/his field seriously, I’ve enjoyed a fair number of business and transformation books. Many of them are good. Some are excellent.
And yet, if I’m honest, from almost every book only one idea truly sticks, one theme that quietly becomes part of my own thinking and philosophy.  Besides simple vague sense of “that was a good book” very often something/one thing stays with me in a way that does not fade, but instead starts shaping how you look at work, organisations, leadership, and change.

Journey to Enterprise Agility was&is an excellent book overall, but what stay with me is this: Transformation should not start with action (but they so very often do start from it).
It should start with forming a common shared world view.
From that emerges the common intent.
And common language follows.
Aligned action comes last.

We almost always start from the wrong end? Most transformations don’t fail because people like to resist change as such. They fail because we rush to action. New structures. New roles. New terms. New ceremonies. New tools. Only later do we wonder why alignment never quite materialises, why progress feels slow, or why teams optimise locally while the bigger picture suffers. When action comes before shared understanding, what follows is not alignment. It is coordinated confusion.

Agile transformations are journeys, not trips? Agile transformations are often treated like projects. “This is how we work from now on.” But “change” and agility do not work that way. There is no final handbook, no permanent end state. Instead, there must be a continuous intent to become better in how we think, communicate, decide and deliver. Agile is not a destination.
It is a journey. And like any meaningful journey, direction matters more than speed.

The sequence that cannot be skipped? The reason the world view intent language action model resonates so deeply with me is simple. It describes a sequence that cannot be bypassed.

Common world view: Without a shared understanding of realities and where we need to head to, people will make rational decisions that lead to irrational outcomes. Different views on who the customer really is, what success means, and where the real constraints lie will inevitably result in misalignment, no matter how well intentioned people are.

Common intent: Alignment is not agreement on plans. It is agreement on intent. When intent is shared, people can make good local decisions without constant coordination. When it is not, even “perfect plans” collapse under pressure.

Common language: Only after intent is widely understood shared does common language start to matter. Frameworks and terminology help, not because they are perfect, but because they enable regular, structured communication with shared meaning. The same words do not automatically mean the same understanding. That understanding must already exist.

Common action: Only now, after having a common world view, intent and language does action make sense. Processes, team structures, roles, metrics and ceremonies belong here, last, not first. Action without shared understanding is noise. Action with shared understanding becomes progress.

Across industries, lack of communication is consistently identified as the number one source of challenges. Agile ways of working and different frameworks can help improving significantly systemic communication. They introduce cadence, shared terminology, and regular interaction. However, resulting improved communication will not fix a transformation that is built on an action-first approach, overlooking the creation of a common world view, intent, and language. When common world view, intent, and language are missing, better communication could merely spreads confusion and resistance faster. “Why did we break something that was working ok?” Frameworks don’t solve this.

Agile transformation is not a team-level exercise, not a “led by agile coaches” initiative, not a process change, and not something that can be delegated away. Leaders must first lead the shaping of a shared world view, help the organisation converge on common intent, and ensure that language truly carries shared meaning.

Only then does action make sense.

Acting too early
Shared meaning not yet in place
Noise travels faster

Work Has More Than Two Shapes — And What We Risk by Not Acknowledging It

Mik Kersten’s Flow Framework https://flowframework.org/ gives a much realistic picture. It distinguishes between four types of work: Features, Defects, Debt and Risk. and I would also add a fifth: Exploration or Innovation. (Innovation and exploration are often assumed to happen “all the time”, as if they were automatic side effects of smart and creative teams. In reality they are not. Exploration requires dedicated time and protected space without the pressure to deliver something immediately. The familiar mantras of “fail fast” or Edison’s idea of finding ten thousand ways that do not work sound inspiring, but they are rarely accepted in day-to-day delivery environments. Exploration challenges assumptions, uncovers alternatives and prevents teams from optimising themselves into a dead end.)

A healthy product organisation should continuously balance these five types (Features, Defects, Debt, Risk, Exploration/Innovation). As Kersten is explaining that balance should not be accidental. It should be intentional, visible and regularly reviewed. Yet base on my experience this almost never happens. And when it does, it rarely sticks.

Why? Because organisations are structurally built to harvest new features. “What are your priorities for the next cycle?” almost never receives the answer: “We’re actually fine with our current feature set — perhaps we should rewrite something properly or explore something entirely new.” Instead, organisations default to creating new things, and only secondarily fixing the mistakes that absolutely must be fixed. Catching up on technical debt slows the pace of new features. Investing in exploration slows the pace of new features. Reducing risk slows the pace of new features. Even a temporary slowdown in feature creation feels dangerous. And so organisations keep pushing for features. They keep celebrating output. They keep adding to the roadmap. And they keep postponing the invisible work — until the invisible becomes unavoidable.

The problem is that debt does not disappear when ignored. Risk does not shrink when left unaddressed. Defects do not become cheaper to fix. And innovation cannot happen in the margins of exhaustion. What happens instead is gradual erosion. Teams lose their ability to move. Codebases/designs resist change. Roadmaps get slower every quarter. Customer issues pile up. Eventually the organisation realises that it no longer controls its own pace. At that point the problem is no longer technical, it is existential. Balancing different types of work will slow down the visible part of delivery. There is no way around that. But the alternative is far more costly: losing competitiveness, one quarter at a time.

Serious, honest and brave organisations and the leaders who run them should aim to make all five types of work visible, set an intentional allocation policy, and review it regularly. The ratios do not need to be perfect. They simply need to be true. The moment we acknowledge that work has more than two shapes is the moment we start building products that can survive, adapt and evolve rather than collapse under the weight of everything we chose not to see.

Perhaps it is also time to broaden the systemic questions we routinely ask. Instead of focusing almost exclusively on “What features do you need next?”, we could ask different, more systemic questions:
“What part of the system would you like to strengthen or make work better?”
“Where would you hope to see more innovation?”
“What risk would you want us to reduce next?”

These questions shift attention from output to capability, from velocity to sustainability and from short-term delivery to long-term competitiveness.

Work has many shapes.
Ignore what hides underneath
and strength turns to sand.

It Is Time to Rediscover Our Responsibility as Listeners in Discussions

Prologue – Why I Begin With Politics in a Business Blog

This time I start with politics, even though this blog is about business, teams, and how organizations work. The reason isn’t political commentary, but the simple fact that political discourse has drifted into a place where organizations should never end up. When listening disappears in politics, shared understanding and the ability to compromise disappear with it. The same risk exists in companies: if we don’t nurture the role of the listener, we risk sliding into unproductive debate, quick counter-attacks, and decisions made without real understanding. That’s why I begin with politics because its failures remind us why a culture of listening is essential for any organization.

Sanna Marin’s 2023 speech to 13,000 students in New York has been itching in my mind for long. “It is time to choose your side,” she said. To me, this time in history is almost the opposite: it is time to understand why someone chooses a different side, not to demand that everyone picks one and to ask how we might find some common ground.

That comment illustrated something important for me about how thin and superficial, self-image driven, political competence has become in many places. The world is growing more complex every year, yet our discussion culture is becoming simpler. Diversity of thought is no longer a richness but a threat to be neutralized.

Polarization is rising and the meaning of compromise is fading.
A functional society is essentially built on compromise, yet we now live in a time where compromise is treated as weakness.

At the same time:

• Social media algorithms push us deeper into our own bubbles. We are not only disagreeing; we are being trained to dislike or even despise those who think differently.
• In political debates, especially during Marin’s era, a culture emerged where nobody truly listened. When one person spoke, another raised a finger not to understand, but to prepare a counterattack.
• Public discussion no longer aims at shared understanding, but at presenting one’s own stance as loudly and confidently as possible.

When “being right” replaces “seeking understanding,” the common ground disappears.

What hands-on work has taught me.
As an experienced hands-on consultant, the kind who doesn’t just tell clients how they should work but actually stays and lives according to the practices I preach, I have learned one of the most important skills of my career:
I almost always have a strong view on how things should be done, but I also understand that there are many valid ways to do things.

When a client makes a decision, I commit to it fully, even if it was not my preferred approach. Joint execution matters more than whose opinion won. To me, this is not weakness; it is maturity. In complex environments, there is rarely only one correct answer.

This same ability to see alternative ways and to accept them seems to have almost disappeared from politics.

We need to rebuild the listener inside us and remember our responsibility as listeners in discussions.
We need to put real effort into not only hearing each other but understanding each other again.

Not in the sense of catching the first word and immediately constructing the next counter-argument, but in genuinely trying to understand why another person thinks differently.

Complex problems will not be solved with quick exchanges, efficiency mantras, or “choose your side” rhetoric. Complexity demands time, space, and respect.

Perhaps the most radical act right now is simply this:
To try to understand.
Not to win.
Not to react.

Ears open, words soften
only when we first listen
can voices truly bloom.

The Number One Problem

Whenever surveys are done, “lack of communication” almost always appears as the number one problem. We have organisational updates, meetings, channels, chats, email threads. It all creates an impression of continuous communication flow. But when you look closely, much of this “communication” is actually very need-driven and often just information sharing in one form or another. Very little of it is truly designed to build a shared view of what matters or to help people find a common direction.

For me, the true strength of agile frameworks like SAFe is that they lead people, both within teams and across functions, to communicate regularly, openly, and with a shared purpose. Leadership and stakeholders align on priorities regularly. Development teams discuss and agree regularly on what is needed next and how much of it they believe they can deliver. Forward-looking happens regularly. Other functions get a reliable, regular opportunity to understand what development is committing to and to adjust their own plans accordingly. Everyone gains regular visibility into how things are actually progressing, and a chance to discuss potential adjustments.

When working on SAFe transformations, I have heard people say, “We’ve always been able to plan regularly and talk to people anywhere in the organisation if we wanted to.”
And the honest follow-up question is: “Have you actually done it, regularly?”
Most often the answer is: “Well… no.”

The possibility alone doesn’t create continuous, unfragmented communication alignment — only regular practice does.

None of the SAFe ceremonies or discussions are driven by individual needs, urgency, or personal ambitions. They are structured to create alignment, they are always bi-directional, and by design each “layer” has the final say on its own topics.

Silent noise everywhere
words flowing, meaning hidden.
True unity born in listening.

When You Describe the Practice, Most of Us Nod.
When You Name the Framework, Many “Panic”.
It’s fascinating, isn’t it?

If you tell people you’re introducing a system where

1. Every 2–3 months, business and leadership review together what’s most important, update priorities, and communicate them clearly to everyone,
2. Every 2–3 months all teams get 1–3 days (depending on the amount of work ahead) to analyse what is being asked from them, ask futher questions and get answers right there and then for leaders, clarify dependencies, raise risks, and finally give their own view on what they can realistically deliver,
3. Leaders respect that estimate, and the resulting plan is shared transparently,
4. Progress is tracked objectively in 1–3 week cycles, without moral labels like “good” or “bad”, just visibility,
5. When things don’t go as planned, it’s nobody’s fault, it’s simply input for collective learning,
6. If real needs change, it’s okay to change the plan because the system serves reality, not the other way around,
7. Everyone ie. developers, product leads, stakeholders, and ideally even customers regularly clarify upcoming work, and nothing starts before understanding is solid enough,
8. When the rhythm stabilizes, everyone knows better where they stand, why they’re doing what they’re doing, and can work sustainably instead of running alongside a bike that never stops. Everyone in the company can plan their own actions better because development teams typically deliver what they promised (as they do not need to & they are nor allowed to overpromise)

…then almost everyone nods and says:
“Sounds great. That’s exactly how we’d like to work.”

But the moment you add:

“We’re implementing Framework X.”

Suddenly the room fills with:

• “That won’t work here.”
• “We don’t need any religious rulebooks.”
• “We’re too unique for that.”

It’s almost comical as the same principles that sounded like common sense five minutes ago become suspicious once they have a name.

Frameworks like SAFe, Scrum, or Kanban are not religions.
They’re just pools of good practices and ways of working, collected from real people who tried to make complex work visible, discussable, and improvable.

They don’t demand faith.
Just curiosity, reflection, and yes, discipline and leadership.

And maybe, just maybe, we don’t need to see only the forest.
Because every tree is unique, valuable, and alive and the framework is simply there to make sure each one gets enough light to grow.

Common sense agrees.
Say “framework”, clouds gather fast
truth lost in the label.

Why Would We Want Anything Other Than Predictability?

(originally written to Wikiagile blog https://www.wikiagile.com/blog/predictability)

Isn’t striving for predictability in everything we do and deliver exactly what we should all want so that the whole organisation can plan and execute as optimally as possible?

But predictability cannot be demanded into existence. Predictability emerges as a consequence when everyone commits to honesty about capacity, constraints, and priorities.

If that honesty is missing and wishful thinking is accepted, overcommitment becomes the norm. When targets are set without understanding/honoring the actual capacity, teams say “yes” because they feel they have no alternative, and plans grow faster than the system can deliver.

In such an environment, predictability is not just absent. It’s impossible.

In the successful transformations I’ve seen, there was a shared choice to face reality:

• Estimates were presented and accepted as they were, even if they exposed uncomfortable limits.
• Saying “no” was treated as a sign of responsibility, not resistance.
• Priorities were reviewed when reality changed, rather than sticking to outdated assumptions.
• Decision-making was transparent and structured, focusing on what could actually be done now.

When this happens, predictability is just one of several benefits:

• Teams can work at a sustainable pace and keep improving.
• Leadership can make decisions based on reliable delivery data.
• Customers and stakeholders can trust the organisation’s commitments.

Predictability is not the goal itself. It is the reward you get when the whole organisation chooses truth over optimism, and discipline over wishful thinking.

And to repeat the starting point, why would you want anything other than predictability from everyone so that everyone can plan their actions according to the most likely outcomes?

We ask for the truth,
yet cheer the sweetest promise
and call it “the plan”.

Truth would guide us well,
but hope sings louder than sense
and schedules obey.

Prioritization: The Forgotten Discipline and Skill – What I’ve Seen Work and Fail

(originally written to Wikiagile blog https://www.wikiagile.com/blog/prioritization-the-forgotten-discipline)

In many transformations I’ve been part of, prioritization has quietly determined whether we moved forward smoothly or got stuck in frustration.

The biggest challenge I’ve seen isn’t typically a “loudest voice” or a political power play. Those happen occasionally, but they’re not at all the main issue. The real persistent challenge is that goals are often set mechanistically for every organisational unit and even for individuals without a tight link or understanding to the actual capacity of the people or teams that must deliver across multiple units.

When you match this real capacity against all the priorities coming from different parts of the organisation, you quickly discover there’s no single authority (~below CEO), and often no agreed process, to prioritise those competing needs for the good of the whole company.

For a somewhat long time, I assumed that creating a single, best possible priority list would naturally be in everyone’s interest and thus balance these competing needs. I thought we could simply put everything important into one ranked list and get on with delivery. But standing in the middle of real transformation work taught me how quickly that list can turn into a logjam.

When this overload happens:
– Trade-offs are avoided because nobody wants to admit their goal might be delayed.
– Teams face constant context-switching, which slows everything down.
– Prioritization becomes meaningless and everything is urgent, but nothing moves fast.
– Constant over-demand erodes faith in the ongoing transformation (“Well, nothing really changed”).

One of the most effective ways I’ve seen to break this deadlock is to use structured decision-making, such as WSJF (Weighted Shortest Job First). The power of WSJF isn’t in the numbers themselves. It is in how it changes the conversation and in who is invited to it.

WSJF works because it:
– Forces us to look at one dimension at a time (business value, time criticality, risk reduction/opportunity enablement, job size).
– Compares relative value instead of debating absolute importance.
– Brings together representatives from as many business areas and functions as possible, ensuring that priorities are evaluated from multiple perspectives.
– Turns “my goal versus your goal” into “let’s look together at which combination gives us the most impact now.”

In this way, it works much like Edward de Bono’s Six Thinking Hats method: instead of mixing all perspectives at once, the group moves step by step, focusing on one lens at a time. This makes discussions calmer, more constructive, and less personal, and often leads to decisions everyone can support, even when their own item doesn’t come first.

Here’s the paradox: most people in organisations already know about these prioritisation tools. But using them is often optional, because there’s a widespread belief that “we already know what’s important.” On top of that, prioritisation rarely follows a fixed cadence. It is assumed to happen “at some point,” somewhere in the process. The result is that prioritisation becomes something vague, informal, and easily overridden by day-to-day pressures.

In the more successful transformations I’ve been part of, structured, inclusive prioritisation conversations were held regularly, with capacity visible to everyone (“This has typically been our throughput”), and leadership strongly supported the outcomes. This didn’t remove tough trade-offs. It made them explicit, and it made them together.

Prioritisation is not about getting your item to the top of the list. It’s about collectively deciding what the system will actually deliver now, and what will wait ….and sticking to those decisions.

And here’s the critical link to transformation: if, planning cycle after planning cycle, teams are overloaded despite their proven throughput and the feedback they give, the whole process becomes demotivating. When that happens, belief in the transformation erodes and with it, the willingness to engage in the very changes needed to make it succeed.

Done well, prioritisation builds clarity, trust, and the credibility of the change itself. Done poorly, it quietly sets the stage for delay, frustration, and a loss of faith in the journey.

Shared skill, shared choices
only collective focus
gives direction weight.

“Retro Comes First” – Why Continuous Improvement Leads To Real Transformation

(originally written to Wikiagile blog https://www.wikiagile.com/blog/retro-comes-first)

This is not meant to be another “you should do it like this” post, but the outcome of a personal retrospective on my past assignments.

Foreword

‍This is not meant to be another “you should do it like this” post, but the outcome of a personal retrospective on my past assignments. I’ve been guilty of thinking “we’ll start proper retros once the organisation is ready,” or squeezing in a rushed session at the end of a Planning Increment just to tick the box. I’ve downplayed timing, skipped follow-up, and failed to build the stamina needed for continuous improvement and thus for real transformation.

I’ve learned the hard way: when you start by undervaluing retros, you make it almost impossible for the organisation to later see them as the engine of change. The baselines you set for timing and follow-up early on are incredibly hard to shift later.

Retros as the Core Drivers of Transformation

Agile transformations often focus on roles, roadmaps, and ceremonies. This is understandable, especially early on when delivery pressure is high. But without consistent, genuine, structured retrospectives, real improvement eventually stalls.

If you’re transforming just to deliver in shorter cycles and not to improve how you work, you’re not truly transforming.

Behavioural change is hard: for individuals (oh boy, don’t I know it … I’d still like to fit into that old wetsuit…), for teams (it’s easier to stick with what we know), and for organisations (where unspoken rules run deep).

Real improvement needs continuous leadership attention and sometimes a firm hand. It’s not enough to ask “Did you hold the retro?” or check throughput metrics. Leaders must help create a system where everyone improves how they work, not just what they deliver. This builds a culture of improvement and a company that’s both more competitive and better to work in.

More Than a Ceremony

Retrospectives aren’t for venting or blaming. They are the mirror and compass of the system. Done well, they reveal systemic issues long before OKRs or KPIs show them. They help teams refocus on how they work, not just what they work on.

Too often, retros are:

• Held late, rushed, or skipped.
• Lacking psychological safety.
• Ignored by leadership.
• Undermined by a belief that change isn’t possible.

Now imagine the opposite:

1. Leaders review retro insights before delivery metrics.
2. Systemic issues are acknowledged and addressed.
3. Teams are expected and supported to try, learn, and improve.
4. Improvement items get top priority, not leftover capacity.

Transformation begins when learning becomes intentional and retros make that learning happen. Without structured reflection and room to experiment, there is no improvement; without improvement, transformation is just a new structure.

Final Remark

Like many personal change attempts, teams and organisations often try to change too much at once and fail. That failure chips away at belief in the ability to change at all.

Confidence grows through small, deliberate steps. Each successful improvement, even if some fail, builds the trust that “we can make changes, we can improve.” That trust unlocks a true continuous improvement culture.

A good Kaizen-style place to start: What is the smallest improvement you could try?

• Everyone on time to meetings?
• Everyone reading materials in advance?

Start small. Keep learning. Stay curious. Stay committed. Ignite the real engine of change.

Speed can blind vision.
Only reflection shapes strength
that lasts beyond now.

SAFe is not just a development model – its true value lies in orchestrating communication

(originally written to Wikiagile blog https://www.wikiagile.com/blog/safe-is-not-just-a-development-model)

While the Scaled Agile Framework (SAFe) is often perceived through its visible elements like ceremonies and plans, its true value lies in its ability to orchestrate and improve communication.

When people are asked what prevents them from succeeding in their work, the answer rarely involves technology, capacity, or competence. More often than not, the biggest challenge mentioned is the lack of communication. Individuals and teams do their best, but often without a regularly shared direction. Decisions are made without an up-to-date overall picture. Priorities shift silently. Information doesn’t flow or it only flows within role-, organizational-, or network-based silos. And even when information does move, it tends to be interpreted differently in different parts of the organization.

The Scaled Agile Framework (SAFe) is, at its core, a response to this challenge. It is designed to create structures for regular, shared communication and the resulting alignment … the kind of foundation that organizations and the people within them need in order to succeed.

In the pressure and constant change of everyday work, it’s only natural that SAFe is often experienced primarily through its most visible elements: ceremonies, plans, workflows, backlogs, and metrics. These are the tangible things. The ones people feel they “have to attend,” measure, and report, and so they often shape how SAFe is perceived. And that perception is often not very positive especially in the early stages of the transformation journey.

The true value of SAFe lies much deeper than ceremonies or plans: it lies in the way it builds a shared rhythm and language of communication across the entire organization. Its built-in practices, structured events, artifacts, and decision points, continuously spark conversations about what truly matters: details, direction, priorities, and continuous improvement. This rhythm doesn’t just increase communication. It balances it. Developers define how and how much. Leaders prioritize and support. And it is in that shared cadence where trust, clarity, and alignment begin to grow. Not through top-down control, but through meaningful, ongoing dialogue.

This rhythm of communication is intentionally built around recurring cycles: Planning Increments bring cadence to strategic communication, while Sprints create heartbeat for everyday dialogue. PI Plannings, demos, and Inspect & Adapt events are not ceremonies for their own sake. They are moments of engagement aiming to foster shared understanding and trust across the board. As a sidenote, along my own journey, I’ve often heard people say, “Well, we’ve always planned, communicated, or reflected on our ways of working together.” But when I’ve asked, “Have you done that regularly and across the organization, based on the most up-to-date overall picture possible?”. The the response has been often silence.

The artifacts of SAFe such as PI Objectives, Actual Value assessments, the Program Predictability Measure (PPM), or the Confidence Vote are not just for reporting. They are platforms for conversation. When teams write their own PI Objectives in plain language, stakeholders can better align with them on what truly needs to be delivered and why. Between planning and evaluation, multiple shared checkpoints allow people to reflect together on the importance, value, and realization of these goals. Actual Value discussions serve as a shared mirror: did we succeed? What helped or hindered us?

The Planning Predictability Measure is not just a retrospective score. It invites broader conversations. Are the needs truly understood? Are our development structures effective? Are there too many dependencies that harm our ability to predict and adapt? These are moments that often trigger re-evaluation of how we organize development work and define deliverables.

One subtle but powerful example of communication facilitation is the Confidence Vote. Its outcome communicates, with a single number, how confident people are in their plan and thus, how confidently stakeholders can build their own follow-up actions. It can serve as an early warning if confidence is low, but more importantly, it opens the door to discussions: what would it take for us to feel more confident? It is not a vote. It is an invitation to dialogue.

The WSJF (Weighted Shortest Job First) prioritization method also exists primarily as a conversation tool. It is not a mechanical formula for what to do next. Instead, it forces a pause – encouraging analytical thought beyond “this feels most important to me.” How much value will it deliver? How urgent is it? How big is the effort? Walking through these questions together enables prioritization decisions that people are more likely to commit to because they helped shape them. WSJF conversations can also produce surprising results, and once again, their value lies in the follow-up discussion they provoke.

The structure of SAFe makes work visible. It produces plans, lists, and metrics. But these are only the tip of the iceberg. What you don’t immediately see – communication, shared understanding, a common vocabulary, and trust – is what makes everything else possible. A real plan doesn’t originate from a slide deck. It emerges from conversations where different perspectives meet, listen, and align.

Lasting change doesn’t come from process. It comes from dialogue. SAFe doesn’t offer a silver bullet, but it forces organizations to talk smart, regularly, and transparently. At its best, it fosters a culture where value is not an assumption but a shared understanding. Goals are not individual desires but mutual commitments. Feedback is not a formality but continuous exchange.

In the end, the success of SAFe does not depend on how well teams can groom a backlog or execute PI Planning. It depends on how well people understand one another, speak about the same things using the same language and want to succeed together.

Structure builds the frame
but conversation brings life;
that is where flow lives.

Agile Transformation: A CxO’s lessons from our journey

(originally written to Wikiagile blog https://www.wikiagile.com/blog/agile-transformation-a-cxos-lessons-from-our-journey)

Becoming agile is not a quick, self-organizing “project”. It is a journey that demands patience, perseverance and active leadership.

‍This is a fictional story from a CxO-level executive about their company’s agile transformation. The story is based on the experiences of a consultant who has worked hands-on in numerous change initiatives. It serves as a summary of key lessons learned, successes, and pitfalls that should be considered when planning a transformation.

When our company set out to embrace agility, we expected a major boost in innovation and productivity. We were convinced that agility was both inevitable and necessary. However, we soon realized that the transformation itself was not simple. Here are the key lessons we learned along the way.

1. Leadership must stay involved

At first, we assumed that agility would emerge naturally if we allocated enough resources and supported it with skilled professionals. We quickly learned that without continuous leadership involvement, the transformation would end up being just a set of new rules without real impact. We had to actively lead the change instead of merely delegating it.

2. Defining value streams requires time and courage

In the beginning, we rushed in our definition of value streams, only to later realize that we had merely reorganized our old structures with new terminology. This prevented real improvements in value creation. We had to take a step back, challenge our assumptions, and redefine our value streams boldly—without being constrained by our existing organizational reality.

3. Focus is essential to make room for the change

Without careful prioritization, agility simply revealed the problems we already had: too many simultaneous initiatives and unclear priorities. We had to make the difficult decisions to stop certain projects in order to create space for real change in the way we work.

4. Structure and discipline are part of freedom

Initially, we gave too much freedom without clear boundaries or a firm, consistent direction. This did not lead to agility but rather to confusion and partial chaos. We learned that agility is not about abandoning structures but continuously improving them. Discipline, guidance, and clear expectations are prerequisites for true flexibility.

5. Organizational politics and emotions must be addressed proactively

Change created uncertainty for many, as it exposed inefficient structures, redundant roles, and outdated responsibilities—even at the leadership level. We realized that without proactive, open discussions, careful planning, and empathetic change management, resistance would significantly slow down progress.

6. Continuous improvement doesn’t happen automatically

Although we officially encouraged continuous improvement, we initially overlooked the fact that it requires structure, clear incentives, and firm expectations. A culture of improvement is of utmost importance in staying competitive and it does not emerge just by hoping for it. It needs dedicated time and systematic support.

7. Leadership must be part of the PDCA cycle

Like many organizations, we initially had a habit of evaluating only final results while neglecting interim progress and continuous learning. Once leadership started actively participating in iterative development cycles, the whole organization became more engaged in listening to customer feedback and adapting quickly.

8. Change happens through emotions

To truly drive agility, we had to create space for people to express their emotions. Rational communication alone was not enough. People needed to voice their concerns and feel that they were part of the transformation. Once we made room for this process, commitment and engagement improved.

In hindsight, starting from this angle would have significantly accelerated everyone’s onboarding to the change.

9. The ROI of communication

When people regularly meet in person across organizational silos, something happens that digital communication alone cannot replicate. Face-to-face discussions, spontaneous idea exchanges, and informal meetings build trust, improve collaboration, and speed up decision-making.

This is not just a matter of efficiency. It is a strategic investment that directly affects an organization’s agility and ability to navigate change. When people know each other better, they better understand how their work fits into the bigger picture, and silos start breaking down naturally. While these intangible benefits are hard to measure using traditional ROI calculations, their impact is seen in daily work: cooperation improves, decisions are made faster, and the organization becomes more responsive to change.

Case: Big Room planning

When we considered expanding the use of Big Room Planning sessions, the cost of the investment raised concerns. It was easy to calculate travel and time expenses, but harder to define how to justify the concrete value of bringing people physically together.

After the first sessions, it became clear that the benefits went far beyond the meeting agenda. Conversations in hallways and informal interactions laid the foundation for an entirely new way of working together. People were not just exchanging information .They were actually getting to know each other, understanding their roles in the bigger picture, and developing stronger collaboration models.

As a result, internal friction decreased, decision-making accelerated, and work coordination became smoother. After each session, our company’s agility improved … not just in terms of processes, but also in how people worked together.

This was not just an operational adjustment but a decision that permanently transformed our organizational culture.

Conclusion: real lessons, lasting change, and a future-proof organization

Becoming agile is not a quick, self-organizing “project”. It is a journey that demands patience, perseverance and active leadership. We had to understand that agility is not just a method but a deep cultural shift across the entire organization. Today, we are more agile than ever, but we recognize that the journey is ongoing. The world keeps moving, and we move with it. Not just to keep up, but because staying in motion keeps us strong and sharp.

Clear rules, steady hands
discipline becomes the space
where real freedom works.

Say No – it is great

(Originally blogged to my personal (now for long passive) blog https://tuomosthoughts.blogspot.com/2017/09/say-no-it-is-great.html)

“No thanks. Not interested.” – what a wonderful & respectful response! Good straight forward fast “No” keeps things rolling and sometimes even accelerating. I admit that this is provocative but I claim that not-replying is one of the very clear sign that the “old good” straight forward  Finnish work culture has collapsed… which we (with about half of size of population of London spread around +300tkm2) in my honest opinion simply cannot afford if we want to stay truly competitive. What do you think why is it so tough to just reply No thanks (especially if you have kind of promised to get back)? Is this causing some sort of systemic malfunction, constant lapses in the general machinery, or does it just not matter? Say it loudly, say it proudly: NO THANKS! 🙂

No takes courage
Silence takes no effort
Systems slow down

(Originally blogged to my personal (now for long passive) bloghttps://tuomosthoughts.blogspot.com/2015/03/mr-ries-nailing-essential.html)

Eric Ries in the lastest Success mag CD with Darren Hardy:

“I am shocked in my travels all over modern capitalism, I have been truly shocked how often people in business are doing stuff but they don’t really know why. Because they were told to do it by their boss.  I heard my boss thought it was a good idea. Well, he or she just knows it  – they are the boss. I don’t know. Or, it’s just how we’ve always done it. We did it last year this way so we continue to do it this way. Just is how it is.
There is an incredible amount of waste that is happening in our modern business.
And you know we have in today’s companies we have our most talented brightest, most creative people on the planet doing this work. And to waste all that time and energy, I think is an appalling appalling waste of human potential.
It’s not really any individual person’s fault. Everyone is doing their job the way they were told. And yet, the net result of sum total is millions and millions and millions of person hours of time and energy being absolutely squandered.
So that’s what we as business leaders, as entrepreneurs, as CEOs, as  managers, as policy makers, that’s we are confronted with. An enormous waste of human potential because the systems that we are building make it so.
And the question we gotta really ask ourselves is that if we learn to be more experimental in the work we do is, there a better way? And I really believe the answer is yes.”

The essential question not only for work life but also even in terms of the survival of the mankind in a world with ever more scarce resources and our limitless appetite for more?

Strip away the noise,
find what truly moves the core —
build upon the rest.

Internal Start-ups as Drivers of Business Renewal

(Originally blogged to Prominda Revolution Oy’s blog August 30 2014)

Over the years, we have been involved in founding and building several internal start-ups within large companies. These have included the development of new types of consumer devices, enterprise devices, enterprise systems, consumer internet services, and technologies – all carried out in start-up mode.

For the past six months or so, we have been analyzing what was done, how the company benefited from these activities, what should be improved, and above all, what value internal start-ups can truly bring to a company. Running them is typically more expensive than so-called “real” start-ups, and their success rate is not necessarily any better – despite the large corporation’s theoretical ability to support them in many ways. So what other value is created beyond the potential of new business?

We have concluded that the benefits of internal start-ups, when done right, go well beyond producing “just” a new product or service. These harder-to-measure benefits relate to the natural renewal of company culture, the building of future competencies, and the adoption of new ways of working.

Internal start-ups help build a company culture that supports renewal. The activity naturally directs people to pay attention to what is happening broadly in the surrounding environment. It fosters a mindset of continuously considering new opportunities, constructively challenging current practices, and responding more quickly to radical changes in the business environment.

Internal start-ups help identify and build the new competencies a company will need in the future. Since start-ups usually aim for future markets with new products and novel business models, they inherently bring in the kinds of skills the company will need going forward.

Internal start-ups also challenge all corporate functions to consider how new initiatives should and can be supported, and how things need to be done differently in the future.

They also have a strong power to engage employees and strengthen commitment. Moving forward with one’s own and shared new ideas often creates an exceptional sense of ownership within the company.

We believe that internal start-ups are among the most powerful ways to challenge and renew a company on a broad scale. Finland now needs, in addition to words, widespread and open-minded creation of the new – especially inside established companies.

Courage is the ground
brave hearts allow new seeds in,
and the old grows new.

How did you end up in such a mess? Two ways. First slowly, then suddenly.

(inspired by Hemingway)

(Originally blogged to Prominda Revolution Oy’s blog September 17th 2014)

Borrowing from Hemingway’s words, I would argue that inefficiency in an organization can accumulate in two ways: first slowly, and then suddenly. Why?

In Finnish companies, the focus has for decades been so heavily on structures that the sense of external threat has widely faded at all levels. The constant concern of winning customers’ hearts has been replaced by the struggle to survive internal tasks and by anxiety about one’s own “position” in a company that is always demanding more efficiency. This is happening despite the fact that the external threat is now greater than ever due to the new competition brought on by global digitalization.

Correcting the course, however, would not actually require much: the key is to fix all structures and reward models that fail to encourage and free everyone to work across silos and to continuously seek the real needs of customers.

The “then suddenly” part of Hemingway’s quote can be avoided by shifting to a start-up mindset – one where people collaborate across silos, help each other, learn together, and renew themselves collectively.

Throughout human history, evolution has favored those capable of cooperation – and it will continue to do so.

First slowly, unnoticed,
then all at once
the mess greets us.

What if Finland’s Social and Health Care Reform Had Been Done Differently?

(Originally blogged to Prominda Revolution Oy’s blog September 25 2014)

In Finland, it was decided to carry out the social and health care (Sote) reform from the inside and together – rather than leaving it to “someone else” to adjust organizational structures and bureaucracy from the top down.

Health and social care workers were taught, motivated, and empowered to think and experiment with how they themselves, and together, could do their work in ways that would continuously serve customer needs better and more efficiently. Numerous administrative experts brought their know-how and holistic perspective into the development and experimentation. We, the customers, were involved not only with our ideas and needs but, above all, as immediate and critical testers of the new service and operating models.

Here and there across Finland, ideas began to pop up about new ways to address our collective Sote needs – and inspired people grouped together to develop and trial these services with customers. They were supported, and failures were treated as shared learning opportunities. The most promising ideas were scaled into wider pilots.

Gradually, and then suddenly, we all realized how the once-declared “inefficient” public sector was actually able to deliver, competitively and continuously, new and better ways to respond to the diverse needs of an aging and increasingly diverse Finland.

And that was not all. The production of Sote services slowly – and then suddenly – became a new Finnish export product. The best new approaches grew into companies of their own, turning service innovation into a source of export revenue. Together, with common sense, we turned decline into prosperity for all.

PS. Finland should have excellent conditions to make Sote services a successful export. We have a high level of competence, strong IT capabilities, a globally low-hierarchy work culture, and a very large national investment already committed to Sote. That should be enough to generate something new worth exporting, shouldn’t it?

Small trials took root
Some grew into something more
Others taught the way