Haute Lumière · The reading

The smartest people in the room keep producing the dullest idea

What gathers in a meeting is not twelve minds but one nervous system, and it decides what can be said before anyone speaks.

She has been at this for an hour. The idea arrived somewhere in the twentieth minute and she has not written it down yet.

THE ROOM

There is a scene so common it has become invisible. A team of capable people gathers around a table, or a grid of video squares, to make something new. They were chosen for expertise, for range, for individual track records. The whiteboard is clean, the sticky notes are fresh, the facilitator is warm and well prepared. Everything is in place. Two hours later the wall holds a set of ideas so safely incremental that everyone leaves carrying a flatness they cannot quite name.

What the room produced could have been produced by one competent person working alone for half an hour. This is not a story about a weak team or a lazy afternoon. It is the ordinary result of gathering, and it repeats across industries, seniorities and continents with a consistency that should have made it a scandal decades ago. The talent was real. The intention was real. The output was thinner than the sum of the people who made it.

Something happens when capable people gather to create that reliably diminishes, rather than multiplies, what they carry individually.

Researchers have a name for the measurable part of this. Production blocking describes the finding that groups generate fewer and less original ideas than the same number of individuals working independently, and it has been replicated across hundreds of studies since the 1950s. It is among the most robust results in organizational psychology and among the most thoroughly ignored. Teams still gather in rooms to brainstorm, decade after decade, on the strength of a belief the evidence has never supported.

The reason the finding gets ignored is that it offers no explanation anyone can act on. Knowing that groups underperform tells a manager nothing about what to do on Tuesday. The explanation turns out to live in three places at once — in what the brain does when other people are watching, in what the body does when a room feels unsafe, and in the fact that people do not merely think differently but construct meaning differently. Each of those is specific. Each of them is addressable.

Begin with the brain, because that is where the first fracture opens. Individual creativity depends on a fluid alternation between two large networks. The Default Mode Network, active in mind-wandering and associative drift, produces the raw material: the odd connection, the unexpected link, the insight that seems to arrive from nowhere. The Executive Control Network evaluates, filters and refines. Good creative work is these two trading places in a rhythm that practised makers learn to trust.

In the presence of others the social environment suppresses the first of them. Threat detection sharpens. The prefrontal cortex moves from exploration to social monitoring, and attention shifts to a different set of questions: what will they make of this, will I look foolish, is this clever enough, is this what the person paying me seems to want. That is not weakness. It is the social brain doing precisely what it evolved to do, and the state it produces is almost exactly the opposite of the state creative divergence requires.

WHAT THE BODY DOES

If you have sat in a meeting where innovation was the agenda and conformity was the outcome, you already know this in the body. Suppression has a texture, a temperature and a weight. There is often a tightening in the chest that accompanies the decision, made below awareness, to withhold the strange idea. There is a narrowing in the throat that makes the words physically harder to produce. Breathing goes shallow, and it goes shallow across the room at roughly the same moment.

This is not metaphor. Stephen Porges's Polyvagal Theory describes the autonomic nervous system as operating in three hierarchical states: ventral vagal, which is safety, connection and creative engagement; sympathetic, which is mobilisation, competition and fight-or-flight; and dorsal vagal, which is shutdown, withdrawal and collapse. Genuine creative collaboration requires the first of these. It requires nervous systems settled enough to play, to take a risk, and to tolerate not yet knowing.

Most rooms designed for innovation push people into one of the other two. Time pressure read as threat, evaluation, hierarchy and the quiet performance of readiness produce sympathetic activation. Longer exposure produces the other thing: the person who is physically present and mentally elsewhere. The facilitator sees heads nodding and records engagement. The bodies in the room are reporting something else entirely, and nobody has been trained to read them.

The most original ideas arrive from the most vulnerable cognitive spaces, which are the first spaces a watched nervous system closes.

This is the part with real consequences. Breakthrough thinking tends to emerge from liminal territory, where established categories break down, where the thinker does not yet know what they think, where a half-formed intuition shimmers at the edge of being sayable. Reaching that territory takes more safety than most rooms supply. So the ideas lost in conference rooms are not the polished proposals people decided to keep to themselves. They are the embryonic ones that never reached the threshold of conscious articulation at all.

That distinction changes what a fix would have to do. If people were choosing to stay quiet, better ground rules would help, and a well-run round-robin would recover most of the loss. They are not choosing. A nervous system reading the room as unsafe closes the channel before anything arrives to be censored, and no amount of encouragement reaches a thought that was never formed. The intervention has to happen earlier, in the state of the room, before the creative question is even asked.

FOUR KINDS

There is a second cause, and conventional practice ignores it entirely. People do not only think differently. They make meaning differently, and the structure through which a person constructs meaning shapes how they approach the creative act. Developmental psychologists call these orders of consciousness, or developmental altitudes. In any reasonably diverse team several are present at once, and they do not innovate the same way.

Read a team through Spiral Dynamics, the framework developed by Clare Graves and elaborated by Don Beck and Christopher Cowan, and a familiar cast appears. The person centred at Blue brings discipline, follow-through, attention to quality, and the moral seriousness to ask whether a thing should be built rather than only whether it can be. Blue innovates by refining within an established frame. Ask this person to ideate wildly and you will produce anxiety rather than ideas.

Orange brings strategic sharpness, competitive drive and a relentless question about whether the thing survives contact with a market. Orange innovates through advantage and measurable improvement, and will generate concepts that are clever and commercially plausible while missing the human and ecological dimensions. Green brings relational sensitivity and insists that innovation serve flourishing rather than extraction, and may resist the decisive narrowing that eventually has to happen. Yellow sees patterns across domains and designs for systemic coherence, and struggles with the patience the others need.

A standard brainstorm is not a neutral instrument. It is an instrument tuned to one altitude, played to a room of four.

Here is the failure. Conventional methods treat all of these people as interchangeable cognitive units: the same prompt, the same sticky note, the same three-minute timer, the same voting protocol. The implicit belief is that creativity is one uniform capacity everyone reaches the same way and that the right technique will open it. Run a standard session with a developmentally diverse team and you have built a process optimised for speed and quantity while quietly disadvantaging everyone who works another way.

Blue's careful contributions are drowned by the pace. Green's concern for who is affected is deferred as not the point right now. Yellow's integrative reframing is too large for a sticky note, gets acknowledged politely, and is lost. Over several such sessions those people stop bringing their real material, the team's creative bandwidth narrows to whatever the method accommodates, and nobody can explain why a deliberately diverse group keeps producing homogeneous work.

Nothing in the room is asking her for anything. That is the condition for the work, not the reward for it.

THE WARM AGREEMENT

Groupthink, identified by Irving Janis in his study of catastrophic policy failures, describes cohesive groups converging early on consensus and suppressing dissent in the service of harmony. What the cognitive account misses is that groupthink is also a physical event. The body learns through long experience that alignment is safer than divergence, and the warmth that accompanies agreement becomes a substitute for the more demanding satisfaction of genuine creative tension.

Watch a room the next time it begins to converge. Bodies relax. Voices soften and synchronise. The energy moves from edgy exploration to comfortable accord, and it feels good, which is exactly what makes it dangerous. The group reads convergence as creative success — we are aligned, we have a direction — when what has actually happened is the early collapse of what was possible.

Status compounds it. Every group carries hierarchies of rank, of perceived expertise, of social confidence, of who is accustomed to being listened to. Higher-status people speak more, are heard more, and have their ideas adopted more often regardless of the quality of those ideas. The most senior person in the room frequently sets the creative direction inside the first few minutes, not by instruction but through a raised eyebrow, a particular question, or an opening frame that narrows the space before anyone has explored it.

These dynamics operate below awareness on both sides. Most leaders sincerely believe they have created an open room. Most participants sincerely believe they are contributing freely. The distance between what people believe about the room and what the room is doing is itself part of the mechanism, because nothing invisible gets addressed.

A group in performance mode cannot innovate. Impressing an audience and discovering something new are different postures of the body.

Then there is the performance trap. When a team knows its output will be presented upward or measured against a metric, a shift occurs that few people notice while it is happening. The work reorients toward producing impressive-looking results rather than pursuing discovery. Teams begin selecting for ideas that will present well rather than ideas that might transform the problem. The question in the room changes quietly, from what is truly possible here to what can we show that makes us look competent.

ATLAS SOLUTIONS

A case makes the architecture visible. Atlas Solutions, a mid-size technology company whose name and details have been altered, assembled twelve of its most capable people for a week-long innovation sprint to design the next generation of its product. Senior engineers, designers, product managers, and two members of the executive team. They hired an experienced facilitation firm and followed a well-regarded method. The room was good. Day One had real energy in it.

By Day Three something had gone quietly wrong. The team had converged on a concept that was, honestly assessed, a modest improvement on the existing product dressed in the language of disruption. Several participants later reported, in confidence, that they had been holding considerably more radical ideas — ideas that questioned the company's business model and the premise of its value proposition. None of those ideas had been spoken aloud.

The reasons, when examined, were the whole architecture compressed into one week. The two executives, despite an honest intention to participate as ordinary team members, had set the direction in the first hour through their reactions to early suggestions. Their visible enthusiasm for some directions and mild scepticism toward others built an invisible corridor, and the team walked down it for four more days without noticing that it had walls.

The most systemically minded person present, an engineer who naturally saw connections across the whole value chain, offered a reframing on Day One that would have opened a far more generative space. It was received politely and then lost. The sticky-note format could not hold it. The schedule could not accommodate the exploration it needed. By Day Two a warm consensus had formed around the safe concept, and one designer who pushed back met a cooling of the room's temperature, read the signal correctly, and stopped.

The sprint did not fail by producing nothing. It failed by consuming the one week in which something could have been found.

The final presentation was polished. Leadership approved the concept. Six months later a competitor launched something close to the idea the systems engineer had tried to raise on the first morning. Atlas had not merely wasted a week; it had spent the only week available on the wrong thing, which is the more expensive failure by some distance. Nobody in that room was incompetent. Nobody in that room could see what was shaping the work from underneath.

FIVE STREAMS

The honest way to read the existing methods is as a set of genuine gifts, each of which cast a new shadow precisely because it lit one area so well. Design Sprints, developed by Jake Knapp at Google Ventures and published in Sprint in 2016, contributed three durable things: time-boxing as creative constraint, structured sequencing across five days, and the Decider — someone with authority to break an impasse so that consensus paralysis does not stall the work.

What the Design Sprint assumes is a cognitive-behavioural human being who needs the right inputs in the right order. The body appears in the schedule as something requiring lunch. The sprint opens with cognitive problem-mapping, which means it opens while social monitoring is at full activation and people are still assessing the room. The structure tells a team what to do. It says nothing about how anyone is showing up, and it ends on Friday with no account of what happens on Monday.

Agile and Scrum brought a deeper shift: complex problems cannot be solved by linear planning, and the most important learning happens through building rather than predicting. The Manifesto puts individuals and interactions over processes and tools, and practice rarely honours it. The standup asks what you did yesterday and what is blocking you. It does not ask what you are afraid to say, or which creative impulse you are suppressing because it does not fit this sprint's goal.

Appreciative Inquiry, out of David Cooperrider's work at Case Western Reserve, established that organizations grow in the direction of what they study, and that beginning from strength generates more creative energy than beginning from deficit. It is neurobiologically shrewd, because appreciation activates the state creative collaboration requires. Its shadow is the shadow of every strength-based approach. Practised carelessly it becomes a container that cannot hold grief, conflict or anger, and the people carrying those quietly withdraw their creative energy from the process.

Theory U, from Otto Scharmer at MIT, is the most ontologically ambitious of the five. It argues that results depend on the quality of awareness from which people operate, and maps a descent from habitual downloading through seeing and sensing to presencing. The phenomenology is beautifully described. What is thinner is the somatic pathway: how a room of executives with activated nervous systems actually arrives at that stillness, and how a facilitator tells real depth from a group performing contemplative quiet.

Every method here is right about what it saw. Each one is silent about the body that was supposed to do it.

The fifth stream is not a method at all. Living Systems Theory — Maturana and Varela on autopoiesis, Prigogine on dissipative structures, Kauffman on self-organization — describes how living systems actually generate novelty: through self-organization rather than control, at the edge of chaos where patterns are loose enough to reorganize and coherent enough to hold their identity, producing emergent properties no component predicted. It reframes the facilitator's work entirely, from directing creativity to tending the conditions under which it organizes itself.

The part that decides everything comes later, on an ordinary morning, when what was found here has to survive a calendar.

BEFORE THE PROBLEM

All of it points at one structural correction, and it is the least popular one to propose: the work does not begin with the problem. A Collaborative Innovation Sprint spends its first fifteen to twenty per cent of time on Attunement, which is the deliberate building of the relational and physical field before that field is asked to hold creative weight. The usual objection arrives immediately, and it deserves a plain answer.

The objection is that this is soft and the real work is the problem. Consider what is actually true of the room in the first minutes. People have arrived from other meetings, other emotional weather, other lives. Some are energised, some are depleted, some are easy with these colleagues and some carry unfinished business with them. The collective nervous system at that moment is a mosaic of different states, most of them performing readiness while scanning for threat.

Build cognitive work on that foundation and the ideas will carry the signature of the states that produced them: cautious, defensive, oriented toward safety rather than risk. Attunement is not a warm-up game, and it is not therapy — it holds a clear boundary against any disclosure a person's system cannot comfortably manage. It is the functional cultivation of the one state from which the rest of the day becomes possible.

It moves in three passes. Self-attunement is the unglamorous act of arriving in one's own body: three conscious breaths, attention redirected from the room to the internal weather, no requirement to change anything that is found there. Sixty seconds of this has measurable effect on parasympathetic activation. The facilitator's own state does more work than the instruction, because one nervous system reads another before it processes any words, and settles in the presence of something settled.

Relational attunement turns the question from how am I to who is here with me. A two-minute paired exchange in which each person offers one honest sentence about what they are carrying — not a status report, a felt state — does two things at once. Each person is witnessed as they actually are rather than as they are performing, and the group's fabric begins to weave through real disclosure. Research on collective intelligence finds that a team's performance depends less on individual scores than on social sensitivity and even turn-taking, which is exactly what this builds.

Field attunement is the subtlest pass: a shift from tracking individuals to sensing the quality of the space between them. For teams new to it this feels unfamiliar, which is expected, because reading a collective field is a skill that develops with use rather than a talent people either have or lack. What matters in a first sprint is the invitation itself — the communication, carried by the structure and by the facilitator's presence, that the space between people is a material the work depends on.

SIX CHANNELS

Divergence takes a quarter to a third of the sprint, and the word to set down at its door is brainstorming. Conventional brainstorming activates exactly one channel: fast, verbal, competitive ideation. It rewards the quick, the confident and the articulate, and it disadvantages the slow, the contemplative and the physically expressive. That is not an inclusion problem in the decorative sense. Different kinds of breakthrough arise from different kinds of intelligence, and a single-channel process yields a single kind of idea.

So this phase runs six channels rather than one. Cognitive-verbal remains, in structured rounds of timed individual writing before any sharing, provocations that invert the premise, analogical prompts that ask what nature would do with the same constraint. Visual-spatial gives dedicated time to sketch mapping, systems diagramming, and arranging physical objects to stand for abstract relations — the channel through which systems-level insight that resists sentences finally becomes visible to everyone.

Somatic-kinesthetic invites the team to walk through the experience they are designing, physically enacting each stage and noticing where the body opens, hesitates, contracts or flows. This regularly surfaces what purely cognitive work misses, because the body's pattern recognition is faster, more integrative and less filtered than deliberate thought. A service flow that reads elegantly on paper can feel exhausting the moment somebody actually moves through it, and the exhaustion is information no diagram was going to provide.

Relational-dialogical acknowledges that some ideas exist in no single mind — they emerge in live exchange, where one half-formed thought catalyses another and the result belongs to neither person and to both. Contemplative-receptive makes room for stillness: solo walks, silence, writing from the felt sense rather than the analytical mind. Ecological-systemic brings the living world in, through biomimicry prompts and the seventh-generation question about what would still be serving life long after the quarter has closed.

Underneath the six channels runs a second correction: developmental calibration. Rather than one exercise for everybody, the phase offers several rounds of different shape. Clear parameters and quality criteria for the person who needs guardrails in order to create. Rapid-fire counted generation for the person who comes alive under a timer. Paired exploration of who is most affected, for the person who creates relationally. Feedback loops and leverage points for the person who needs the challenge to match the complexity they already perceive.

Invite four kinds of intelligence with one exercise and three of them will politely decline.

Midway through, the facilitator introduces the Shadow Sprint Protocol. Each person writes, anonymously, on a card: the idea I am not saying because. The cards are collected, shuffled and read aloud without attribution, and each is received with curiosity rather than judgement — so there is something here we are not saying because it touches a project somebody senior is fond of, that is useful to know, let us put that territory back on the table. There is usually nervous laughter, then relief, then a surge. The group's actual bandwidth, rather than the portion it felt safe to use, becomes available to the work.

THE BODY DECIDES

Convergence is where most processes quietly fail, and not by skipping the phase. They execute it mechanically. Everybody receives a fixed number of dots, places them on preferred ideas, and the ideas with the most dots proceed. It has the look of democracy. What actually occurs is a compound of status effects, in which people watch where the senior person's dots land before placing their own; conformity, in which clusters attract further clusters regardless of merit; and anchoring, in which ideas positioned first or at eye level collect disproportionate attention.

The result is convergence on what is popular rather than what is profound, and in group settings those two frequently point in different directions. The alternative is not a cleverer voting scheme. It is a three-stage practice that lets alignment emerge rather than be tallied.

First, the full divergent harvest goes up, visible at once and deliberately unsorted. The mess is the point: it is the edge of chaos that living systems theory identifies as the zone of maximum creative potential, and organising it too early collapses the possibility it holds. Then comes somatic sensing. Participants walk slowly along the wall attending to bodily response rather than evaluating — what opens something in the chest, what produces a felt no, what produces a signal they cannot yet put into words.

This leans on established ground. The somatic marker hypothesis, from Antonio Damasio's work, holds that emotional and bodily signals carry real evaluative information and often arrive before rational analysis has finished. So participants mark responses rather than cast votes: one colour for yes, one for not yet, one for something is here and I cannot articulate it. That third marker is usually the most interesting thing on the wall, because it points at the liminal place where something genuinely new is trying to form.

Third, the group reads patterns rather than analysing them. The question is not what should be the priority but what is the field telling us, where is the energy gathering, what wants to come together. This routinely surprises everybody, the facilitator included: ideas that looked unrelated turn out to be facets of one insight, a concept with few votes and many uncertain markers turns out to be the seed, two apparently opposed ideas turn out to be a polarity more powerful held together than resolved.

The discipline it takes is tolerance for not yet knowing, held longer than anyone in the room finds comfortable. That discomfort is not a sign of failure; it is what emergence feels like from inside. Collapse the tension early and the group lands on something familiar. The opposite error is refusing to narrow at all in the name of including everyone, and the answer to that one is that convergence is selection rather than rejection. What is not chosen becomes the compost of a later idea.

Prototyping follows, and its governing principle is that the prototype is not the idea but a conversation with it. Speed over polish, because a polished artefact invites praise and a rough one invites the truth. Alongside material prototyping the sprint uses somatic prototyping, in which people physically enact the innovation and arrange themselves in space to feel what a proposed structure is like to occupy; narrative prototyping, in which a specific person has a specific experience on a specific Tuesday morning; and systems prototyping, which maps how the thing will ripple outward before anyone builds it.

MONDAY MORNING

The last phase is the one almost every process treats as paperwork, and it is the one that decides whether anything survives. The pattern is familiar to anyone who has run a good workshop: real energy, genuine prototypes, useful insight, and then Monday, when people return to their desks and the work dies quietly in the gap between the room and the rest of the organization.

That gap is not accidental. The conditions that made the sprint work — dedicated time, freedom from mail, permission to be uncertain, a room of people paying attention to each other — do not exist in ordinary operating life. The sprint built a temporary zone in which creative work was possible. Integration is the deliberate construction of bridges out of that zone, and there are four of them.

Translation re-expresses the insight in the language the organization actually uses — priorities, resources, timelines, stakeholders — without losing what made it worth having. This is harder than it sounds, because the richest insights are precisely the ones that resist existing categories. A translation nobody outside the room can follow is useless. A translation everybody understands that has quietly discarded the innovation is worse than useless, because it produces the appearance of change while delivering the previous thing in new packaging.

Sponsorship names out loud who will carry the work through the political landscape: who champions it, who must be persuaded, what resources it needs, what will obstruct it. Living follow-through sets specific deliverables, owners, dates, and a date to reconvene — a plan with one unusual element, which is a short scheduled check-in where the team reconnects for ten minutes to sense how the thing is developing rather than only to report on its status.

An insight that cannot be re-entered is a memory. An insight with a body attached is a capacity.

The fourth bridge has no conventional equivalent. Before the sprint closes, each person finds a brief somatic anchor: a gesture, a posture, a breath pattern that encodes the felt sense of what they discovered. This is not symbolism. It rests on state-dependent memory — what is accessed in a particular state is more easily retrieved when that state is re-entered. People who have an anchor report reaching the sprint's intelligence again in ordinary weeks, before a difficult meeting, in a moment of decision. The work becomes portable rather than stranded in the room that produced it.

That is the argument in full. The innovation crisis is not a shortage of creative people and not a deficit of good methods. It is that the methods address what a room does and leave untouched what a room is: the nervous systems in it, the different structures through which its members build meaning, the material nobody is saying, and the question of whether what gets made will serve anything living. Those four are specific, and being specific, they can be designed for. The book is how.

Free to read

Free to read, and free to hear. Every chapter of every book in this house, and every narration of it, is open to anybody. No account, no card, nothing to cancel.

What is in it

• Collaborative Innovation Sprints™open this in the house search

A nervous system scanning for threat cannot also wander. It chooses safety, every time, without asking.

Most ideas are not withheld in meetings. They are never formed.

Agreement releases the same warmth as discovery and costs a great deal less.

You cannot brief a room into safety. You can only build it, and it takes twenty minutes.

Keep looking

Every phrase on this page opens into the house search. The shelf holds The Developmental Canon and six other shelves, and the reading is free.