If you ask ten people what are the top five things you can do with your hands, I bet hardly any of them will tell you that your hands are excellent for speaking. Unless they are Italians. And even then, most of us are so used to using our hands to communicate that we all take it for granted, sometimes becoming a beloved foreign laughing stock. But talking with your hands does wonders for the speaker and the audience, which is why I include hand gesturing while speaking among the things that matter if you want to train your intuition.

However gesturing is still not necessarily popular at work. “Try not to be so passionate” is something I have heard multiple times, most memorably from a colleague whose own family had arrived in Britain a generation earlier, from a part of Asia that carries the same passionate reputation Italians do. But they have not developed hand-gesture as much as Italians have.

My take is that he had worked very hard at keeping still. I see the same thing in my two daughters, born to immigrant parents, working very hard on anything they subconsciously associate to be stereotypically British, from composure to lack of passion to their accents. They find it something to hold on to in a desperate need to fit in. But the struggles of immigrant parents and first generation children and their quest to be accepted doesn’t belong to this series so I will leave it alone.

After researching the benefits of hand gesturing associated with fostering intuitive signalling in ourselves and others, what I wanted to find out was why Italians spoke with their hands and the British didn’t. Some online references to busy marketplaces and the variety of dialects present in Italy felt slightly superficial and unconvincing to me, so I dug deeper into history.

Why Italians speak with their hands

I started from a little refresh on Italian history, which seemed like the obvious place for my investigation. The top line is that Italy has been one country for the last 165 years. Before, it was a series of city states with their own dialects, with roughly half of this time spent as a republic, as the king was sent away and the monarchy abolished in 1946. There is a role that monarchy has to play with hand-gesture and we will come to it later.

During the years, from the Etruscans to the Romans, Italy’s trading partners and the fact that it had been invaded for centuries by foreign populations, Greeks in the south, Normans in Sicily, Spanish and Austrians later, made a developed hand signing language a necessity and not a nice to have. It most likely also allowed people to speak behind the back of the latest ruler in a way that was harder to trace.

Why the British learned to keep still

England had a lot of varied languages too. From Welsh to Gaelic to Old Norse. The difference is that the Tudors did a brilliant job since 1500 to eradicate, and social police, any alternative form of communication they could not understand, and hence control. The “King’s English” became the only language accepted for law, trade and administration.

They went even further and built stigma and social status associated with “the stiff upper lip” as the only acceptable way to be. Still and controlled was good, opposite to the “uncivilised” gesturing Italians used to communicate, even accounted for by British travellers coming back from Italy in the sixteenth and seventeenth century, as uncivilised and vulgar. Heavily relying on verbal language only became the thing classy British had to comply with.

Where gesture became a full language

Italians have 250 codified gestures uniting us as a nation, and almost all other Mediterranean populations from Spanish to Israelis use gestures profusely, even though they never quite got to the fame Italians achieved as hand talkers. 250 codified gestures is nothing in comparison to the thousands of unique signs the Native American tribes devised to coordinate large buffalo hunts across millions of square kilometres, negotiate borders, form alliances and hold gatherings with thousands without uttering one word.

The Australian Aboriginal peoples developed similar skills, also related to hunting coordination, but also to complicated social rituals, including mourning practices under which women specifically were forbidden to speak, sometimes for months. Their hands remained their main tool for communication with the rest of the family and society.

What gesture does for you and for your audience

Not only is it a good exercise in intuition to second guess what the other might be meaning with their hands, but gesturing also helps you get your point across in the fastest of ways.

A study from 2025 conducted by UBC Sauder School of Business found that speaking with your hands makes you more persuasive. They analysed more than 2,000 TED talks using AI multimodal models, and the speakers who used their hands in a way that visually supported the point they were trying to make came across as much more competent than the others. The audience grasps their message faster because hand gesturing functions as an effective tool to aid the brains of the audience in understanding, and remembering, the point.

There is also something worth noticing in how they did it. Multimodal models now read gesture. What used to be invisible is becoming measurable with AI.

On the speaker side, talking with your hands provides cognitive offloading as well as aiding your mental file finding, supporting you in finding in your mental lexicon the most appropriate word to use. Susan Goldin-Meadow shows clearly that gesture is an integral part of the way we speak, and that restricting the hands measurably degrades the effectiveness of how well people explain and reason.

Audiences remember better when they can subconsciously use two parts of their brain to decode and remember, as explained by Allan Paivio’s dual coding theory, the one associated with words and the one associated with images.

Research from the Max Planck Institute for Psycholinguistics found that listeners were twenty per cent more likely to hear and remember a word correctly when a beat gesture fell on the stressed syllable. What is even more interesting is that when the gesture fell on the wrong syllable, listeners were forty per cent more likely to mishear the word altogether. So your hands have to be used coherently as they are an integral part of the signal, and out of time they can corrupt it.

Where AI fits and what to try this week

Try Voice mode. Speaking a problem aloud while moving your hands, then bringing the resolved version to the prompt, restores what typing removes.

Another thing worth trying is showing yourself from the waist up in your next video call. Your hands in sight, so that your audience gets a chance to benefit from seeing them moving as people will find following you easier. Framing that cuts you off at the shoulders removes the entire second channel, and the Max Planck finding suggests that is not neutral, it is lossy.

Stiff upper lip or open hands, the choice is yours

Do you want to learn how to talk with your hands well, or follow the old fashioned “composed” stiff upper lip etiquette that in my view was mainly effective from the Tudors onwards to keep population better under control with obsolete social conventions?

I have a feeling that intuition has been gradually and systematically trained out of us, and although I do not intend to go into the educational polemic or speculating on the political reasons why this might have been the case, I am feeling proud and lucky to be also Italian today, with dual citizenship, as I believe talking with my hands has helped me and my audiences a lot during the years.

I have a little treasure in my hands, like the little shell, a precious gift from my daughter, resting in my palm captured in the image I used for this piece, and you are welcome to start practicing too. Your intuition and audiences will benefit from it.

Next week

Next week, what changes the moment you say the problem out loud. Why people solve things halfway through explaining them to someone who has not said a word, and the strange but genuinely useful gift I received from my team in 2008.

If you are not yet following Intuition Edge, you can do that here:

Share

During Covid, Chiara used to call me and sometimes, with her usual kindness, ask me to get out in my garden, rain allowing, and to start pruning roses. It did not take me long to realise that every time she did that, she had something important to brainstorm together, something that often had different options all worth being considered, something that often demanded not to stop at the first couple of obvious answers. I do not think I ever told Chiara that I really did not like roses. I followed her nudges religiously anyway, weather allowing, and went into the garden, scissors in hand, pruning methodically everything else but roses as I was talking to her on the phone.

I did ask her once why she wanted me to get into the garden when talking to her and she said she wanted the whole of me, not distracted by other screens or incoming messages, and I seemed to always be sharper when strolling or pruning my plants.

A slow, repetitive, almost mindless movement which requires just enough focus to occupy my hands and completely free the part of my brain that was talking to Chiara. Many options were discussed and created, attention focused and alternatives found in that garden.

It took me a very long time to work out that she was following her intuition in driving me where she subconsciously knew without knowing, that by encouraging me to busy myself outdoors on a slow and repetitive task such as gardening, she was in effect encouraging me to create the condition to think better and getting a better brainstorming buddy for herself. Chiara and I had worked together for over fifteen years by then. What she was doing for me, it turns out, works for almost everyone.

Why slow movement sharpens your thinking

I already wrote about the power of the pause for creative problem solving and what a fern can do to your attention, and I think both pieces still apply here, but there is something more. How you move changes how you think, the very nature and quality of your thoughts, which is why movement belongs in any serious conversation about how to train your intuition.

Slow, repetitive movement improves the way your prefrontal cortex, the area of your brain sitting right inside your forehead, which is the area responsible for memory, flexibility, planning and reasoning, or in short our executive network, performs, as illustrated by Hannah Critchlow in The 21st Century Brain. Gardening is one type of slow, repetitive movement, yet simply taking a gentle stroll anywhere might work just as well.

We know it is down to how we move in its own right thanks to an experiment conducted by Marily Oppezzo and Daniel Schwartz at Stanford, where one of the conditions involved an indoor treadmill facing a blank wall. So no trees, no view, no restoration, in effect nothing at all to look at. It produced results close to walking outdoors.

Why walking helped me lead better

I subconsciously and purely intuitively encouraged my creative teams to take a walk in the park or a stroll through the Tate Modern or any other London exhibition when they were stuck for ideas, or when we did not have the luxury of time because we were up against a tight deadline on a high-profile, competitive pitch in my advertising or digital product days. I literally sent them for a walk.

It seemed a bit crazy at the time, as others might have been more inclined to chain them to their desks, yet the results were more often than not outstanding and allowed me to build a reputation as a great closer, and the pitches I led and the stories I built supported by the right evidence were the most effective, possibly coupled with a cohesive team in a room that hardly ever did an all-nighter but had had a very productive creative day strolling through the Serpentine in Hyde Park. It worked for all of us. Years later I proposed walk and talk meetings in a company I was leading, suggesting members of the team choose a minimum of two meetings a week that could be held while walking. We were never as disciplined about it as I would have liked, yet with the spread of remote working, and for digital and technical product teams who spend far too long in front of a screen, it would have been additive and healthy.

What I find really interesting is that walking barely affects convergent thinking, the kind where a problem has one correct answer that you need to hunt for. Oppezzo and Schwartz published in the Journal of Experimental Psychology: Learning, Memory and Cognition in 2014 an experiment where eighty-one per cent of participants improved on the open-ended task and only twenty-three per cent on the one with a single right answer. So walk while you are generating, come back, and we can all sit down when we have to decide.

What happens if you go faster for longer

Annie Murphy Paul, in The Extended Mind, one of my favourite books I read last year, asks you to imagine movement as a bell curve. On the one hand you have slow and repetitive, such as walking, gardening or peeling potatoes, helping you raise your executive performance to an optimum state. On the other hand, you have long, effortful, sustained movement, like running a marathon. Slow and repetitive make that part of your brain associated with executive thinking work better, long, harder and sustained shut down that very same area.

The forever Nobel favourite Murakami explains beautifully how it has worked for him. He talks about “the void”, the state of absence of thoughts, achieved through running a marathon. Reading him, I am not sure how liberating running is for him rather than a sort of compulsion yet nonetheless the state achieved is described incredibly well.

Arne Dietrich has proposed a fascinating hypothesis, although with caution,He offers the idea of transient hypofrontality, where higher prefrontal functions are temporarily turned down so that the body can allocate all its resources to the motor and sensory systems. An active state while fully awake, where the part of your brain associated with your implicit systems gets freed up from structured and aware frontal control as your executive brain is quietened.

An interesting physiological argument for why long effort gives you access to what you already know but found hard to explain, which is not a bad description of intuition when you think about it. Nothing usable turns up while you are still running obviously, since you need all your energy not to think but to run, yet as you stop intuitive signals have perhaps a better chance to surface, most likely somewhere inconvenient like the shower.

Two different creative superpowers

Thankfully you do not have to become a marathoner to enter this space. You can opt for swimming lengths, cycling, rowing, walking hard uphill for an hour. It just needs to be long, hard, sustained and kind of boring.

If walking widens the field while you still hold it, and sustained effort removes your hold altogether, then I believe we can actively and deliberately use these practices to improve our intuition and how we can deal with it once it has surfaced.

I imagine being a sculptor. Walking gives me a useful set of tools for me to pick the right one to shape as the task in hand. Running gives me the block of clay to mold. The first gives you the instruments whilst the second gives you something to use them on.

Where AI fits into your thinking

Assisted intelligence is brilliant at converging as it hands you a resolved answer, confident, plausible and already pretty narrowed.

Assisted intelligence needs assisted judgement, and part of that judgement is knowing how wide the field was before you let the machine tempt you to close it too soon.

Forty minutes of walking is the most effective way to widen it first, so that what you bring to the prompt is a set of options rather than a request for one. The same time spent running will temporarily send that executive function to sleep, so that you can see if anything more useful pops up on the other side of it.

This week

Take one decision that is still open, and rather than “think about it”, perhaps “walk about it”. Or why not move a couple of virtual meetings to walk and talk? You will most likely come back with options rather than a conclusion, and then you can sit down and decide.

It turns out it does not work just for me and my slightly ADHD-ish wiring.

Thank you Chiara for making me discover the real power beyond tidying up a messy garden.

Next week

I will share the power of talking with your hands. As an Italian I always took it as national heritage, and all the historical explanations I had been given for why Italians gesture so much turn out to be nowhere near as interesting as what it actually does for your thinking. Until then.

If you are not yet following Intuition Edge, you can do that here: substack.com/@intuitionedge / Published every Monday.

Share

Last week I recommended putting four heads together to solve a problem, and in today’s story there are five in the room, so I believe I need to explain why. In Laszlo Bock’s Work Rules!, four was the ideal number of people to assess the same candidate in an interview, which is a much narrower problem than a strategy workshop. In a workshop each participant usually carries a different part of the problem.

So the optimum number in a room varies according to the problem at hand and the type of company you are solving it for. I believe it was Jeff Bezos who first introduced pizza in relation to management. According to him you had to limit the attendees of a meeting to a size small enough to be fed with two pizzas, roughly five to ten, for a bunch of good reasons such as focus, accountability and speed. I agree with him, even though I think the pizza reference only works in the States. If you are Italian, and Roman as I am, one pizza feeds one person by design because it is a different pizza, so please do not run a ten-person meeting in Rome on classic Roman pizza takeaway or they will starve.

Jokes aside, I did some research and found some interesting numbers floating around. Hackman and Vidmar’s 1970 study in Sociometry points at 4.6 as the ideal number. They also found that size made very little difference to how well the groups actually performed. What the 4.6 really tells you is the point at which people stop complaining about how many, or how few, of them there are. Fewer than four and you do not have enough genuinely different perspectives in the room to be worth debating. But if you get past six, as Hackman himself concluded years later in Leading Teams, the session starts being about the people in it rather than about the work.

It is important to say that the experiment was conducted with students on a fifteen-minute task, so hardly comparable to a strategy workshop. I do still think size matters in a workshop, though what matters far more is who you put in the room, what you tell them and when, and whether the conversation is genuinely happening at group level rather than between two people while everyone else watches.

Five Skilled Practitioners And An AI

The five I am thinking of were sitting round a table with me and an algorithm, and the session had not started, because we were still deciding how to brief the machine. That turned out to be the session. One of them had been feeding it narrow questions and getting narrow answers, having never considered that the framing was the constraint. Another had assumed that brand-sensitive work needed extensive training and careful curation, and had barely used it there at all. A third had not known it could hold an entire technical build plan and help think through the order things needed to happen in. A fourth shared a prompting approach that visibly changed how two people in the room would work from that day on, and the fifth walked everyone through her own method for checking whether an output was valid, which the group adopted on the spot. They had all been using it alone and almost intimately, and building something with it together made that visible in a way solitary use never could.

You Are Not Alone

In my last piece I wrote about the dangers of using AI in an intimate one to one relationship. There are many ways to use AI well, and the most important of them is making sure there is a collective driving towards better, with different biases, the machine’s included, coming together in a well structured session. That is the thing worth curating. You are never alone, even in a CEO job, which is notoriously deemed a lonely one. Well curated co-creation with your team, and now with the AI, is the way forward to better decision making.

The Gains

Once you have assembled the right team, the job is to make sure collective intuition is well harnessed and managed. The team’s individual intuitive signals, captured live and verbalised through a shared language, can be collected in a process that creates a safe space and real progression on the task at hand. I call that process SPACE, and these are the five benefits I keep seeing when a room uses it in a structured workshop with corporate teams:

• Momentum. People align and decide faster as a team, thanks to a shared language, a process, a safe space and a record of why.

• Fewer bad decisions from unchecked AI.

• Risks identified earlier, because someone can now safely express what they are sensing, having been given the language for it.

• Better accuracy and differentiation. If everyone uses the same tools and writes similar prompts you get a sea of sameness, and curated difference is what counters it.

• Clear ownership at board and senior leadership level. People point at the model and accountability moves away from the person. AI is not a cover. Accountability stays with the team using the tool well.

The best prompt needs a good deal of exchange, just as the best AI output needs stress-testing collectively, using each participant's intuitive signalling alongside their professional biases and whatever other data is relevant.

The Collective Can Do What No Individual Can

So if biases are a plus, and problem solving in 2026 is the art of humans and machines coming together, how do we learn to write the best prompts?

Together.

The best prompts are the ones crafted collectively. Keeping the team small and curated around solving a problem from different perspectives, rather than filling a room with people who will tend to agree because they share the same experience, is half the job.

The other half is letting an experienced external moderator guide the conversation, since their biases and viewpoints are usually different from those of everyone employed by the same company.

Disagree About The Work, Never About The Person

When relationships suffer the work suffers with them, and never more so than on complex decisions. Here is the hack. Avoid arguing about who does what and instead argue about how the thing should be done. De Wit, Greer and Jehn reviewed 116 studies across nearly nine thousand groups and found that task conflict only helps when relationship stay healthy.

It comes down to sentences like "You always over-engineer this" or "That's a classic marketing thing to say", which immediately become personal. Instead try phrases like "I read that output differently, and here is where" or "What would have to be true for that to work?", which are about the work. And never feel afraid to ask the team to hold collective judgement open for ten minutes in a world that is pushing us to decide super fast, often prematurely.

Why AI Makes Arguing Well Easier

ChatGPT and Copilot do not care about being promoted or about pleasing one person over another, and they have no reputation or position to defend. Once the prompt and the context have been decided together, nobody in the room solely owns the output either.

In one of the sessions I ran for a global brand I work with, the most interesting finding was accidental. Challenging Claude's output turned out to be easier than openly disagreeing with a colleague, which made it a very effective catalyst for surfacing what the room really thought. It was task conflict with the relationship conflict taken out of it, and a group of people who have just argued with a machine find it considerably easier to argue constructively with each other. Everyone in the session left feeling more necessary than ever, at a time when employees are sabotaging AI implementation initiatives for fear of losing their jobs.

What This Means For You This Week

If you take one thing from this, make it the prompt. Write it with other people, out loud, and expect that to take longer than you think it should. Most of the arguments worth having in a session turn out to be arguments about what you are actually asking, and they surface while you are still deciding what the machine needs to know, do and where it has no saying.

Get someone in from outside if you possibly can. I would say that, since it is what I do for a living, though the reason is genuinely that everybody employed by the same company has been marinating in the same assumptions, culture, category dynamics, and is key to have someone in that room who is alien to that.

Then use SPACE to say out loud what you are sensing, and put it on the table next to the numbers rather than underneath them.

Have fun with it and tell me what worked and what did not. I hope this turns out to be the year AI stops being used as a solo confidante.

What’s Next

Back to nature, and to the way we move, and to the practices that make your own signal more likely to surface in the first place.

If you are not yet following Intuition Edge, you can do that here: substack.com/@intuitionedge

Published every Monday.

Subscribe now

Please allow me to start with a little disclaimer. The sheep you are looking at has no role to play for about six hundred words or so, but when her moment comes it is a key one, which is why I chose her as the header. Please bear with me.

The inspiration for the last piece, this one, and the one coming next all started while I was at MAD//Fest last month, when someone in the AI and Innovation stage audience raised a hand and asked a question that I felt was a little under answered and not given the full response it deserved. As I was the Chair of that stage, and time keeping was among my duties, I did not linger and I let the conversation move on rather than pausing, yet the question kept bugging me.

More or less, the question was this. I am advising a client, he is making ads independently with Claude, they are not that great, and he does not want to listen to me even though he hired me as the expert to guide him.

Your question carried a lot of depth, and yes, it made me think about the IKEA effect I covered in last week’s piece, which answers the first part of it. Why does your client think his work is better than it is. The second part, which is just as interesting, went unanswered. Why does he not listen to you, given that he hired you to advise him.

Maybe he was a victim of the IKEA effect. Or maybe he thought the consultant was biased, which is the very reason she deserved to be hired in the first place.

Subscribe now

Bias As An Asset

Nobody is independent or unbiased, and that is a good thing when it is actively used, as I covered in The Filter Nobody Told You About. We are the product of our experiences and we look at the world through the tinted glasses of wherever we focus our attention, as well as the goals we are given or set ourselves inside the organisation we serve. Our wishes, desires and fears often run us, and while a good deal of self discipline can at least make us more aware of them, nothing beats a well curated team that pulls in different directions because it is differently biased. AI is one member of that team. The secret sauce is how you curate who is in the room, and the space and the rules you give them, and it, to interact.

And there is nothing worse in my view than group bias. As Kahneman, Sibony and Sunstein knowledgeably point out, while averaging opinions removes the scatter between them, it does nothing at all to a bias they share. If you ask ten people to guess the weight of a sheep they will come up with random guesses, and the average will be more or less right. Now show them the sheep at the top of this piece, photographed head on, in low evening light, standing rather closer to the lens than the rest of the flock and every one of them will most likely guess high, so the average of ten high guesses will simply be a higher overall guess. In their book Noise they use a cow, but you get the gist.

This is what you get when five CMOs in the same sector, all reading the same newsletters and all measured on the same quarterly numbers, get to debate something. They will most likely agree with each other, at best bringing some tricks and tools and shared personal experience as added value to the conversation, and real innovation is unlikely to happen because they will tend to confirm what they more or less already knew. AI is the worst of the lot on this, because what it hands back is roughly the consensus of everything already written on the subject.

Your Secret Sauce Is You

Even maths proves it, as Scott Page showed that a group’s collective prediction error equals the average individual error minus the diversity of their estimates, which he calls the Diversity Prediction Theorem. In layman’s terms, collective accuracy rises because people think differently, even when no single person in the group improves their own reasoning or expertise.

His caveat is that this only holds inside a shared base of relevant competence, and outside it what you have is noise rather than signal. Or in other words, you would not invite a skateboarder into a brainstorm aimed at working out whether the Higgs boson can explain the matter-antimatter asymmetry in the universe. Unless the skateboarder is also a physicist, of course. All explained in Your Best Career Hack Is The Collective.

So Where Does That Leave You

Try to assess with some level of lucidity where your competence really lies. You can use my diagram here, which was inspired by something Warren Buffett popularised, so trust that it is at least half good. Then ask a few friends or colleagues to tell you what they think of your core abilities. I tend to pick four, two who have known me for a long time and two who are relatively recent acquaintances, and more on why later. All four share enough language and domain competence to understand what they are looking at, while bringing slightly adjacent and different skillsets, so that each of them sees something the others do not.

Beauty is in the eye of the beholder, and it is rather nice to find out what people see in you that you could not see yourself, provided you encourage radical candour from them and receive what comes back with radical open-mindedness. The test for assessing your own baseline on that is in When Did You Last Change Your Mind On Purpose?.

Then curate the group itself, because knowing where its boundary sits is what tells you whether its disagreement is going to be additive.

How Few, And In What Order

Do we need to keep asking an infinite number of relevant opinions? As I said, I tend to ask no more than four people, and I was thrilled to discover much later in life that there might be some science behind it. Bock, in Work Rules!, describes how in Google’s interview process four interviewers get it right about eighty-six per cent of the time, and every interviewer after that adds roughly one per cent of precision. Four kept apart beats sixteen more, and that is assuming the four reach the decision as a committee. Anything beyond it is not worth the time of the interviewers, let alone the interviewee.

What This Means For You This Week

Count how many of the people you routinely ask are wrong in the same direction as you. Then, next time you are unsure, ask four colleagues and friends. AI can be one of them, and not the main one, and certainly not the only one. Although without AI I would not have been able to generate the extra grass and flock that made my picture, originally shot in portrait, work as a landscape header on Substack.

What’s Next

This was heading for three thousand words, which felt like a lot to ask of anybody in August, so I have kept it to a size you can read between the beach and the second iced coffee of the day. Next week, what I did promise last: how I use AI as one member of a curated team and how to make our biases an asset in practice rather than in theory.

Published every Monday.

Yes, I did go hunting for a bug, or a scarrafone, in every room of my garden in Puglia, and I finally found this one. Yes, I am keen to keep using my own photography as much as I can throughout my posts.

But the inspiration for this piece did not come from a bug, it came from a conversation with a very smart friend who told me recently that she had taken to running her thinking past AI because it was less biased than the people around her, and at that point I could not refrain from preaching at her more vehemently than a Jehovah's Witness on a doorstep.

If you have been reading me you will know by now that I find machines incredibly useful, able to synthesise, retrieve, create and model from known data at a volume and speed none of us had before. They have even helped me become clearer in my own strategies and voice, through the slow process of rejecting the alternatives they offered and working out, each time, exactly what it was that I did not like. They have saved me hundreds of hours on board packs, research and analysis. And they are incredibly biased, as is everyone else around you, which is a subject big enough that I am giving it a piece of its own next week.

AI As Your Counsellor

I paused for a meaningful while as I was choosing the word counsellor in the section title just now. I could have picked advisor, or tool, or even companion. Yet I settled on counsellor, which in my view carries technical and emotional value in equal measure, and a counsellor is confidential. Nobody else is in the room, nothing is written down where anyone can see it, and no colleague ever finds out what you asked.

We have been using AI in an intimate, private way. Perhaps when you asked ChatGPT “what does EBIDTA mean” rather than asking your Finance Director. Or when you spent twenty minutes of back and forth seeking emotional reassurance that no, your boss is not cross with you because of the way she phrased that email, and yes, you do sound a little defensive in the way you have formulated your reply, and no, you are not overreacting. Or, worst of all, when you asked whether your idea was any good before sharing it with a colleague, in fear of somebody repeating it or, worse still, passing it off as their own. Your team might never get to see that idea.

Why We Started Asking A Machine Instead

The reasons range from fear, to shame to a real or perceived lack of emotional safety, in a world that is going slightly wonky and where AI is genuinely taking jobs.

Challenger, Gray and Christmas attribute 101,743 job cuts in the first half of this year to AI alone, and a Harris Poll survey for Express Employment Professionals found last year that seventy-three per cent of job seekers believe no job is safe regardless of how well the person does it, with sixty-seven per cent saying they used to feel stable in their careers and no longer do. has just written a whole book, I Eat the Stars, about how to live fully in a world that is quietly coming apart, and it is top of my summer reading list. Whether or not you accept her diagnosis, the feeling she describes is the one people are carrying into the office every morning. We run to AI for a great many of the answers, and in most cases that works perfectly well. In others it costs us an opportunity we never find out about, or at worst it starts to erode the very capabilities we are relying on, because we do not integrate what we received with our own intuition or, most importantly, with the collective, which is where the next piece goes.

This one is to share a couple of dynamics I have watched form in rooms over the last year or so that deserve noting, so that we can move forward positively and with intent.

1. Hurray, I Can Do This With AI Now And I Don’t Need You (Designer, Engineer, Lawyer) Anymore

I have come up with a hypothesis that took inspiration to the Buffett and Munger circle of competence, just that I imagine the circle having more rings outside it. Towards the centre there is what you know really really well, or your area of expertise if you like, and as you move out it is what you are less of an expert of, until you reach the boundaries and meet vaguer and vaguer notions of the topics at hand. The size of your inner circle of competence does not matter. Knowing where its boundary is very much does. Sometimes good friends and colleagues can help you assess it.

How much it matters that you cannot tell what good looks like depends entirely on the context and the risk carried with it. As a way of direct example, I made the diagram above myself, in about the time it would have taken me to brief somebody. A designer would look at it and see a dozen choices I did not know I was making, and would probably have made better ones. I looked at it and saw four rings doing the job, which tells you something about where the boundary of my own circle sits, and I decided the twenty minutes I saved were better spent on the words.

On a more serious note, two friends of mine negotiated redundancy settlements last year. Both of them ran the process primarily with AI. One came out of it with the most impressive outcome I have ever seen anybody achieve in such a settlement, and the other missed a big opportunity after almost two decades of employment.

They were using much the same tool. One of them had spent years inside commercial negotiations and had exited a couple of companies already, so she could feel where the advice was going thin. The other had never done any of it before.

Telling apart intuition from noise is something I covered in [The Two Conditions That Decide If Your Intuition Is Right]. Your signal is only worth trusting in a domain regular enough to have patterns in it, and only if you have had honest feedback on your judgements often enough to have been corrected. AI output can only be judged if you have enough experience to discern if it is good enough or not, and when it perhaps doesn’t matter.

2. I Am More Right Than Ever

The second dynamic is more annoying and more damaging. AI allows you to barricade yourself inside your own conviction, supported by data and a history trail that it was simply not possible to keep track of before. You arrive with a position, you ask a question shaped by that position, and back comes a beautifully organised case for the thing you already believed, with footnotes you can now annoy everybody else with.

I wrote about the antidote in When Did You Last Change Your Mind On Purpose?, and I am well aware that recommending open-mindedness is easy while actually start practicing can be quite hard. You can start from here. Before you go looking for support, write down what would have to be true for you to be wrong, and go asking for that instead.

3. Thirty Seconds To Get An Answer = Thirty Seconds To Decide

The tempo of the tool becomes the tempo you feel you have to match with your judgement. An answer that would have taken a week to assemble now arrives while you are still pondering whether you should have added that extra bit of context to the question, and the gap where the signal used to form quietly disappears.

I have written before about what a possum taught me about thinking, and the short version is that the pause was doing something. You can keep the speed, by all means, and yet not feel pressured into deciding today.

4. I Made This, And I Made It Before I Could

Which brings me to the scarrafone. In Naples they say “ogni scarrafone è bello a mamma soja”, every cockroach is beautiful to its own mother, and I have never found a better description of what happens when somebody produces an artefact they could not have produced alone a year ago.

There is proper science underneath the proverb. Norton, Mochon and Ariely gave it a name in 2012, the IKEA effect, and across four studies of people assembling flat-pack boxes, folding origami and building Lego, participants valued their own amateurish creations at roughly what they would pay for an expert’s work, and fully expected everybody else to agree with them. The detail that fascinated me is the boundary condition: the effect only appears when the task is completed successfully, and it disappears entirely when the creation fails or is left unfinished.

Sit with that for a moment in the context of a machine that completes every single task you give it, on the first attempt, most times in a presentable format. We are building a pretty efficient manufacturing process for the IKEA effect. You are proud of having made something that was out of your reach a year ago, and so you should be. What that pride does not tell you is whether the thing is any good.

Couple that with my first point and you are building some real blind spots.

So What To Do

Start with an honest assessment of your own competencies and capabilities. Have you done enough of this particular thing, and received enough active feedback on it, to know when good enough is genuinely good enough and when it is not?

Then ask the trusted people around you, which sounds obvious and is something we do less and less of, because the whole appeal of the AI counsellor was that nobody had to know.

More on the personal mini board, and how to assemble one worth having, is coming. In the meantime, be radically open minded about whatever comes back at you, and let it shape the next iteration rather than the defence of the last one.

What’s Next

How I use AI as one member of a curated team, how to make our biases an asset, and why the team that argues well beats the individual with the best prompt.

Send this to somebody who has recently made something they are very proud of, and be gentle about it. Every scarrafone deserves at least one honest reader, which brings me to my own. This piece has had two beta readers, my husband and Claude, one of whom is contractually devoted to me and the other of which is trained to tell me I am great. So please do tell me if it is part rubbish. I don’t like blind spots.

Published every Monday.

At least I seem to be on the right side of an average.

Substack has launched Pangram, an AI powered tool that claims to detect, with 99% accuracy, whether a piece of text was written by a human or by AI. It sorts your writing into three buckets: human, ChatGPT-esque, or Claude-sounding.

I ran it on three random paragraphs from a handful of my articles. It scored me 100% human the first time, 100% AI the second, and 100% human again on the third. By then I had rather lost interest, realised I was only part synthetic, and decided I was fine with that.

On one hand, I fully understood ’s sentiment behind the war on slop. On the other hand, like Reddit and a good many others, I found it quietly ironic that the weapon of choice against AI was another AI powered tool.

Optimisation will be OK; Creativity remains controversial

Most leaders, some employees and definitely most investors, see the value in AI for optimisation, even where that value is still latent and has not quite materialised. Don’t worry, it will. Creativity is where the agreement ends and the spectrum runs wide.

And the debate could not be more concrete. Coca-Cola ran an AI-generated Christmas ad two years running, doubling down the second time despite the first having been widely panned and the industry kept talking about it until summer. Cannes Lions, saw award entries drop by more than a quarter, after a 2025 scandal in which a Grand Prix was revoked for AI-manipulated footage, that made mandatory AI disclosure and detection tools. A festival losing seven thousand entries the moment it asked people to declare their AI use tells you something about how much was quietly going unspoken.

Is it acceptable to write with AI, or only to edit with it? To use it for production, or to let it reach the concept? It has been a long time since I saw this much genuine debate about anything.

My question underneath it is this: what is it right to use AI for, in what capacity, and most importantly, who gets to decide?

AI made writing better, and samier

AI made a great deal of creativity online better and made it more of the same. Writers who have made a living from the craft for years reacted strongly against text they suspected was machine-written.

In January I wrote a piece for MAD//Fest titled “Would you know, and care, if ChatGPT wrote your birthday card?” It caused a small riot. I did not write it for the clicks, but to find out whether the sentiment would beat the craft. Craft won, thirty to seventy.

Does that mean I am not disappointed when I click an intriguing title and the first line opens with “it’s not X, it’s Y”? It does not. Does it mean I stop reading if the piece covers something I am genuinely interested in? Also no. That article was prompted, shaped and approved by an author who, one hopes, cares about what they choose to publish.

What AI actually did for me

Without AI, you would not be reading this. Not because I wrote it with AI, but because I would never have found the confidence to press publish, or the time to research the old school way.

As a child, writing carried me through the school system with good grades despite being an undiagnosed dyslexic with, I suspect, a few ADHD traits tagging along too. Writing, philosophy and theoretical physics pulled my average up. Algebra and anything too rigidly linear dragged it down. I left Italy at seventeen and picked up French and English along the way, working up to an executive role in London. My English writing became competent but very technical, very corporate, prose in a PowerPoint bullet. With AI I found the courage to attempt non-fiction again, and I liked the result.

AI never writes my work yet I write my text in a moltitude of ways. Sometimes at my laptop in the garden, as today. Other times I write chunks by hand on A4, knowing that four and a half pages of my handwriting come to roughly a thousand words, then photograph them, drop them into Claude, and ask it to suggest structural improvements and type them up.

And here is where it became genuinely useful, and deeply meaningful. Using AI forced me to ask radical questions about my own voice. Not what it was, as I have always known that, but which of the things I learned along the way I actually wanted to keep and which I wanted to reject. I learned to write in Italy, where the em-dash was never explained to me, so I never knew it existed until AI-generated content arrived. I understand now what it does, and how useful it can be yet I actively choose not to use it, because it is not part of how I learned to write. That choice is intuition in the service of authenticity, applied to my own craft, the same trained instinct this whole series is about, pointed inward.

It has produced some odd moments. When I pushed back on being edited into short, staccato sentences, Claude told me I had “a more European style,” and I confess I didn’t know how to feel about that. What did that mean, exactly? That declining to write in clipped, anxious little fragments made me somehow more European? And the best came from a senior Italian friend, also working in London, to whom I returned his own newsletter with a gentle note about a spelling mistake. Do not worry, he said, he was seeding those in, because people had started telling him his newsletters read as though they were AI-generated - which I thought they were anyway. All those years as a dyslexic Italian chasing decent English spelling (or decent spelling in any language, for that matter) to climb up the marketing ladder, and now the errors were the alibi. Something in me staged a small rebellion.

The tools prefer human work anyway

I do not fully understand the use of the detector, beyond earning twenty dollars and up from those who really need to know whether a machine had a hand in a piece. To me, an original idea, even if only so-so written, even with the occasional em-dash AI giveaway, is worth reading.

Lean on AI too heavily and you pay for it in reach. Graphite now puts primarily AI-generated articles at roughly half of everything newly published in English, and the machines do not appear to enjoy reading their own work, since 82% of what ChatGPT and Perplexity actually cite is human-written.

Poor find-ability is a price I can live with. What I am less sure about is software telling me to change how I write, or to lean less on the tools that get my ideas out.

There may be edge cases where I change my mind yet so far, the clear winner looks like Pangram’s shareholders to me.

The real risk

Not quite what the detectors are looking for but what my very good friend wisely said: “The problem isn’t people using AI and passing it off as their own work. It’s people not using AI and sounding like they have, because that’s what they think good writing is.”

That is the thing to watch. AI is changing how we write en masse, and, faster and more strangely, how we speak. I have never noticed as sharply as in the last few months how many people around me have started saying “landed” or “this is the whole point.”

So. If you were an average five and AI brings you to a six, so be it. Keep your ideas your own, and stay true to the point of view you want to bring to the world, even if the words are AI-assisted. If you were a seven or above, like Nick, you have nothing to fear, because there will always be a market for you.

The founders of Pangram will make their money, since a witch hunt has always been good business for the hunters.

But I see something better in this spiky, uneven world where the natural and the synthetic sit side by side, like the strange green ball I photographed on a concrete pavement on my way home.

If you care about what you write, and the intent and the ideas are yours, use AI to take your words to a homogeneous six. It is better than what the internet used to be. And if your ideas are interesting, I will still read you.

What’s next

Out of my own head and into the room. How to use AI as one member of a curated team, how to run sessions where human intuition and machine output sharpen each other rather than blur into the same six, and why the team that argues well beats the individual with the best prompt.

If you found this useful, share it with the most splendid human writer you know. Or with the one who was bad and quietly became rather good, yet now reads like ChatGPT. Or with the one seeding spelling mistakes to prove they are still real, if you happen to know one. I suspect you do.

For two weeks I have been telling you to put the screen down and go and find your possum, or talk to a fern. There is plenty of scientific evidence that both of those activities do wonders for your creativity and judgement. There is also plenty of room to explore the leap forward in creativity and judgement that AI can deliver.

The machines have made the case for trained intuition far better than I ever could, and considerably faster than I was managing on my own.

I said it on stage last week as a chaired the AI and Innovation stage at MAD//Fest in London and I will repeat it here. Had I not kept reacting to ChatGPT and Claude producing strategies, copy and creative solutions that sounded persuasive and did not sit well with me, I would never have clocked the urgency of training our intuition alongside the optimal use of these tools, which grow more powerful by the month.

Five things I have heard in meetings.

I put these on a slide at MAD//Fest and watched a room of senior people laugh in the particular way that tells you they have done at least three of them. I could probably have written twenty more. I would love your own in the comments, though I suspect people remain somewhat private about their conversations with these tools, and about what they do with the output that is quietly informing their decisions at work.

The Delegation.

Twenty-five minutes spent crafting the perfect prompt in order to save ten. Mine often started, “I now want you to act as the most competent strategist in the world…”

The Expert.

Ninety-five per cent excellent and correct, with one statistic slightly too convenient, which made me look like a fool in front of a colleague who had thankfully decided to check it.

The Authority.

“Because Claude said so,” delivered with a straight face, as though that settled the matter. I promise you this happened, in a recent workshop, from a senior person I respect a great deal, at a global brand I was advising.

The Thanks.

Should I thank the algorithm? If you are British, thank yous and sorrys are national insurance, so a server farm somewhere is hopefully appreciating your kindness. Jokes and nationalities aside, you are not alone in the confusion. Research cited by Arizona State University in The Lancet Child & Adolescent Health found that forty-two per cent of American teenagers have used AI chatbots for friendship, and nineteen per cent for romantic relationships. Susana Ortega, a seventeen-year-old on the research team, points out that “with artificial intelligence, it’s programmed to like you.” But I am clearly getting distracted by the fascinating sociocultural rabbit hole here.

Friday, 5.55pm.

It felt off but it had the right flow and the charts looked good, so you shipped it. Monday at 9.02am, you knew exactly why it had felt off. We did this long before AI arrived, because tired is tired. What has changed is that the confidence, speed and completeness of the answers we now get make cognitive overload, and cognitive offloading with it, a more serious business, and our bias, which I suspect is becoming collective, towards speed over quality far more evident.

Every one of those hands something over. The work itself is entirely fine to hand over. The judgement about the work is another matter, and these moments are so frictionless that you can pass through all five in a single afternoon without once noticing you have made a decision.

Why this got urgent rather than merely interesting.

At MAD//Fest, Stjepan Zelić, CEO of Hypefy, asked a great question. Since we are now flooded with AI-generated output, has creative work online got better or worse?

The answer, once he walked us through it, was obvious. The median human sits at around a five out of ten across creative ability. AI hauled everyone below that line up to a six. Writing, image selection, design decisions, the whole lot. The bad output vanished almost overnight and the average rose from a five to a clear six. The people already above a six stayed exactly where they were, because the machine had nothing to teach them and possibly a great deal to learn from them.

So the work got better and far more homogeneous. Doshi and Hauser found the same thing under laboratory conditions in Science Advances: AI-assisted stories were judged more creative and better written, and were measurably more like one another than the human-only ones.

Yet the machines don’t seem to like their own work, which in itself is ironic as Graphite now puts primarily AI-generated articles at roughly half of everything newly published in English yet around 82 per cent of what ChatGPT and Perplexity actually cite is human-written.

The machines are already casting their vote with their citations as they prefer the human stuff.

What that does to the value of judgement.

If everything worth creating now lives in the space between the invisible six and the ten, then where you stand becomes your edge. Access to the tool is universal. How you use it, how you react to its outputs, and how you weigh your own signals alongside the machine’s, that is your assisted judgement.

Which puts the weight on the least respected capability we have, the one we have been steadily trained out of in favour of analytical outputs that can be fully explained. I wrote my manifesto about it on day one of this publication, so I will spare you the repeat.

Intuition has always been fundamental, to keeping us safe and to making anything genuinely new. What AI changed, and what I am grateful for, is the urgency. When the median was a five and arrived slowly, you could get by on process. Now the median is a six and arrives in four seconds, and the only thing separating your work from everyone else’s is the judgement and the taste you bring to it.

Assisted intelligence needs assisted judgement.

Let your intelligence be assisted by your trained intuition used alongside data and AI - even better when exercised collectively. Use AI to hold more variables than you can, to draft faster than you can, to surface patterns in data too large to see and use your intuitive signal to guide you on how to make the most out of it with better judgment.

But if the intelligence is assisted, the judgement has to be assisted too, and by different things. Your intuition needs to be methodically trained rather than merely trusted, and integrated with other good data, AI, and other humans. Imagine AI as one member of a team you have curated, and not the head.

Three years ago I would have told you that training your intuition was a good idea, in roughly the way that learning to write well or taking up swimming is a good idea. Now I would say it is the load-bearing skill, and that everything else being sold to you as an AI capability sits on top of it and falls over without it.

What this means for you this week.

As we learn that cognitive debt is a real problem, identified by MIT in people using ChatGPT, and that opinion shifts are definitely happening, subtly, there is however another read on what AI gives us.

It can enhance your capabilities, inspire you and, as it has me, help you know yourself better by surfacing questions you were not consciously asking, and reactions worth listening to. I have had plenty of revelations about my own use of language this way, though that is a story for another piece.

So take the five moments above. Work out honestly which one you did most recently, and how you reacted. Then, as you decide what context to give a prompt or what to do with what comes back, notice the flinch, and start treating your intuition as the hard, tangible capability it is, in your own judgement and your team’s.

I am still very much figuring out the optimal balance of delegation, which shifts constantly as the tools improve, as I widen the range of tasks I am comfortable handing over, and as my own abilities evolve alongside them.

If you would like to explore how to structure sessions where a curated human team uses its intuitive signalling alongside AI to solve a specific problem, get in touch and I will show you a few ways to do it in context.

What’s next.

The creative half of this argument, which I am rather looking forward to. How AI genuinely extends what you can make once you drive it with a clear sense of your own voice, why I formalised my tone of voice this year through some deliberate choices about who I am, and what that did to the quality of what came back. And why I have largely stopped using it one to one, and started using it as one member of a curated team.

If you found this useful, share it with someone who has said “because the AI said so” in a meeting this month.

If you are not yet following Intuition Edge, you can do that here: substack.com/@intuitionedge

Published every Monday.

Share

It took less than three weeks to get my “you were right” out of my committed city-slicker husband, after I finally convinced him to move the family to the countryside.

Our life in London had been a privileged one. We lived a minute from one of the city’s greatest parks, and our flat had a couple of terraces I had turned into small jungles. We moved to a house at the back of a forest, with majestic beech trees. He was unsure at first and won over in days – the branches did it.

Last week I wrote about the possum who taught me to leave my desk, and about the incubation effect, the reason a stuck mind solves a problem best when you stop staring at it.

Then I spent three enjoyable days chairing the AI and Innovation stage at MAD//Fest London, where I opened with a keynote on training your intuition to make better decisions in the age of AI, and where a speaker I will come back to, gave a brilliant talk based on the book that runs underneath this piece.

Whilst last week was about the pause, this one focuses on what to look at once you have taken it.

We are all just trying to get home.

We build gardens for the same reason we drift towards windows. It is longing for the conditions we evolved in and left behind.

Annie Murphy Paul’s The Extended Mind is one of the most elegant books I have read on this subject in years, and it makes an observation I have not been able to shake. She describes landowners in the parched American southwest irrigating their properties until they resemble lush grassland, and gardeners in Japan pruning their trees so the boughs take on the spreading shape of an East African canopy.

These people are not following independent trends or stylistic choices but are chasing what the biologist Gordon Orians calls the “ghosts of environments past”.

We long to return to, or in this case recreate, the places that once kept us safe. Which is why, when we are stressed or agitated or stuck, we so often find ourselves walking towards a park, a hill or the sea without quite deciding to. We seek solace in nature, and nature repays the courtesy of the visit by deeply regulating us.

What the research actually shows.

Gregory Bratman at Stanford found that a ninety-minute walk in nature reduced rumination, that grinding loop of negative self-focused thought, in a way an equivalent walk beside traffic did not.

Marc Berman at the University of Chicago found that people who walked for under an hour through a leafy park performed 20% better on a test of working memory, compared to those who did the same walk alongside busy city roads.

People who have spent time in nature also catch more errors when proofreading, and answer faster and more accurately under cognitive pressure.

Which brings me to fractals.

The geometry your nervous system likes.

There are many fractals in nature, patterns that repeat themselves at every scale, so that the smallest part echoes the whole, a snowflake, a pinecone, a shell. A fern is the clearest example you will ever find, which is why one sits at the top of this piece. Yet so are coastlines, branching trees, clouds, and so the list goes on.

Mathematicians measure fractal complexity on a scale from one to three. Most of nature’s fractals sit between 1.3 and 1.5, and it turns out that human brains prefer precisely that band, intricate enough to hold the eye, simple enough to require no concentration. The physicist Richard Taylor has spent years demonstrating that looking at fractals in this range measurably lowers stress and restores tired attention.

No surprise, then, that when the adult colouring craze took hold a decade or so ago, several million stressed adults, given every possible subject, chose to spend their evenings shading in repeating botanical patterns.

Whether this was the fractals working on them or just the soothing effect of staying inside the lines, the research has not settled. It has, however, settled that they did not choose to colour in spreadsheets.

And this is what your screen cannot do. A screen is a machine for extracting attention, however restful its contents seem. A fern gives it back.

You do not need a forest.

I now live at the edge of one, which makes me the wrong person to tell you that you do not need it. Yet I can promise you that this mechanism does not care about acreage, just about having somewhere gentle for your eyes to land.

A balcony, a single street tree, a city park you cut through on the way to the station. They all do the trick. Or a plant sitting close enough to your desk that when you look up from the screen your eyes land on something that returns your attention rather than spending it, that does it too, for about five pounds.

Where AI does not belong, again.

That speaker was Alec, and he came at all of this from the opposite direction.

I had argued from the stage that intuition is a capacity you have to train, and that we will need assisted judgement alongside assisted intelligence. A few hours later Alec asked what we lose when the machine starts doing our thinking for us. Cognitive offloading, synthetic empathy, the quiet erosion of human judgement.

We found we had read many of the same books, The Extended Mind among them, and had arrived in the same place from different directions. He was asking what depletes our judgement as I had started from what restores it. The answers turn out to be many of the same, and I suspect I will be seeing a good deal more of Alec.

So here is the provocation. Talk to a plant more often than you talk to Claude. Not because the tools are useless, they are extraordinary, and I will spend the coming weeks writing about them. But when your attention is spent and your thinking has gone flat, the answer is not another tab or another prompt. It is to look at something that costs you nothing to look at, until the tank refills and you can think again.

What this means for you this week.

Put something green within the eyeline of where you work, for the moment when your eyes lift from the screen and need somewhere to rest that is not another screen.

Then, once a day, let them arrive there and leave them for a full minute. No podcast, no phone, no purpose. Just be.

What’s next?

Next week I turn back to AI properly. I heard too many good speakers and saw too many good tools at MAD//Fest not to pay some of them their due.

If you found this useful, share it with someone whose desk has nothing living on it.

If you are not yet following Intuition Edge, you can do that here: substack.com/@intuitionedge

Published every Monday.

Share

In 2000 I moved from Rome to Melbourne. Inconveniently, in the middle of my master's degree and a good part-time job, I became convinced within a few months that I had lost my intelligence.

It mattered more than it might have. Work was easy to come by in Australia then, and I had somehow secured a huge job I was truly under-qualified for. I got it on the strength of being hard-working, keen, and conveniently fluent in a couple of European languages, while my English was, let us say, “in progress.” I needed every wit I had to keep it.

I had come from a country where you could still smoke at your desk, in a meeting, in a restaurant, to one where stepping outside to smoke had already become the done thing. And just like that, with the mere act of quitting, all my good ideas seemed to have deserted me. It was obvious: the cigarettes had been making me clever, and quitting had broken the spell. I was doomed.

Except I wasn’t. It just took me embarrassingly long to understand what was actually going on.

My intelligence has not gone with the cigarettes, it had simply removed the ten small excuses a day that used to send me out of the building, away from my desk, away from the problem, to stand under the gum trees doing nothing of any consequence. My eyes would settle on a possum who lived in the branches, not studying him but simply resting somewhere while my mind wandered off entirely. He asked nothing of me. Thank you, fluffy little thing with the huge eyes. And while the two of us sat in that companionable uselessness, the back of my mind quietly got on with the work the front of it had been failing at. I would wander back to my desk and the answer would be there, as though it had let itself in while I was out.

It was never the nicotine but the break. Which meant cigarettes had, as far as I could now tell, no redeeming feature whatsoever. So I did what any newly enlightened person does: I became a born-again anti-smoker of the most insufferable kind, the zeal of the recently converted, someone who has Seen The Light and would like a word with you about it.

The science behind the break.

This turns out to be one of the sturdier findings in the study of how we solve problems.

Psychologists call it the incubation effect: set a problem down and your mind keeps working it in the background while your attention is elsewhere.

The strongest version comes with two conditions, and I find both quietly reassuring.

First, the break should be gently undemanding, a walk, a wander, a stare into the middle distance, rather than another taxing task, because it is the loose, drifting quality of an idle mind that does the work. Second, it helps to know you will come back to the problem. There is a subtle self-reassurance in this, as you are not abandoning the problem or giving up on it, but simply handing it to the quieter part of your mind and stepping out of its way. You are giving your subconscious room to work, and your intuition more chance to show up. No one puts this better than Rick Rubin in his beautifully written The Creative Act: A Way of Being, a long argument for getting out of your own way and letting the work arrive.

I wrote about the deeper mechanism, opportunistic assimilation, when it took me five years to act on a doctor’s advice. The possum version is the same thing, minus the daddy issues.

What did the possum teach me?

That is really all the possum was teaching me, by example. He did nothing, required nothing, produced nothing visible, and was entirely, magnificently fine. It took me a quarter of a century, and a fair amount of reading, to understand he had the right idea.

Where AI does not belong.

A note, since I am about to spend three days chairing the AI and Innovation stage at MAD//Fest. The coming weeks here will be full of AI. This piece, deliberately, is not, and that is the actual tip buried inside it. Part of using these tools well is knowing where they do not belong, and a stuck mind reaching for a better answer is precisely one of those places. Do not open another tab. Leave the screen at the desk. Go and find your possum.

A small confession.

I have not transcended my past. I am still, somewhere underneath, one of you, a smoker who has simply managed to stay quit. The craving never fully leaves; it just learns some manners. So I will not preach, though if you light a Marlboro near me I will quietly, nostalgically, breathe it in. I will only say that the best thing the cigarettes ever gave me turned out to be the one thing I could keep without them. The walk away.

What’s next.

Next week, why where you take that break matters as much as taking it, and why a fern on a windowsill can do what a forest does.

If you are not yet following Intuition Edge, you can do that here: substack.com/@intuitionedge

Published every Monday.

Share

If you have read a few of my pieces, you already know where the next paragraph is going, so skim ahead with my blessing. But do stay, because this piece is practical, simple, and easy to keep in your head when you actually need it.

Your body reaches a verdict before your mind has finished reading the evidence. You walk out of a meeting that went well on paper and something sits wrong in your chest. A deal everyone is celebrating leaves you faintly cold or oddly uncomfortable. A hire who ticks every box gives you a flicker of doubt you cannot quite articulate. All of it is information being served to you, and most of us have been trained to ignore it in favour of the spreadsheet and the slide deck.

Last week I set out the two conditions that decide whether a signal is worth trusting: a world regular enough to hold a real pattern, and enough honest feedback to have learned it. This week is about what to do once a signal arrives. How to read a red flag from a green light, and how to act when your gut and your data fall out with each other.

The green light I followed.

Years ago, working with Fiat, I had a hunch. Give people a simple way to see their environmental impact as they drove, in real time, and they would change how they drove, gentler on the planet and on their wallet, saving fuel just as a recession was closing in. The hunch did not come from a dashboard, and it was certainly not mine alone. It grew out of a series of conversations with Luis, an enlightened client who had a hunch of his own: that Fiat’s Blue&Me platform, the Microsoft system built into a large share of cars already on the road, and used mainly for safety, could do much more. I still remember him talking to me about music and in-car entertainment. It was 2007.

It became eco:Drive, and it went on to win a Grand Prix at Cannes Lions, a CNET award and a Gold at the Clios, then rolled out across seventeen markets. It is not about the trophies, but about the fact that the intuition arrived before the evidence. The timing was right, people were ready, and the data caught up to confirm what the signal already knew. That is a green light: gut and evidence pointing the same way.

Eco:Drive taught me to trust the signal. Which is exactly how, a few years later, I came to ignore one.

The red flag I overrode.

Then there was Opnuu, my peer-to-peer fashion lending marketplace. It came from the same place as Eco:Drive: a read on the circular economy and its promise of something better for the planet and the wallet at once. And it came from a kinship with another Livia, a Brazilian who loved sharing her frocks as much as I did. I sat in a meeting with a well-known VC who went through the pitch and found no fault in the model or the execution. The closed beta results were strong yet, disappointingly, something bothered him enough to keep his chequebook firmly shut.

‘Why is there no directly comparable peer-to-peer fashion platform in the US?’ he asked.

I didn’t know why, but I was steaming. I decided he was being narrow, mistaking caution for territory bias. Americans were already using Airbnb, TaskRabbit and Bag Borrow or Steal without blinking. Close enough, surely, to count as proof that Europe was ready too.

What in fact was happening was that he was telling me something I did not want to hear. If the US, usually ahead of Europe on peer-to-peer adoption, had not produced this model yet, the market might not be ready. I dropped the warning, raised what I could, and pushed on. I had to shut it down a couple of years later for lack of funding.

Here is the part that took me years to make peace with. My instinct was right: peer-to-peer fashion rental became a real category, and ByRotation built a thriving business almost ten years later on the idea I had pitched. I had read the signal correctly and misjudged the main thing that decides who wins: whether the world is ready to receive it yet.

So this is something else I learned about intuition. It can see the world as it will be, not as it is, so even when the vision is right, timing always wins.

The grid I built so I would never do it again.

Nearly a year ago Julian Gomez invited me to guest lecture at Hult International Business School. Preparing for it forced me to articulate something I had been using for years without ever structuring it. A simple grid for putting intuition back where it belongs, beside the data rather than beneath it.

I first shared an early version of it in a piece for MAD//Fest late last year. This is the 2.0, the version with everything I have learned since, including the one change that mattered most.

Two axes. What does your intuition say, yes or no. What does the data say, yes or no. Four outcomes.

Gut yes, data yes. Proceed. This is alignment, and it is rarer than you would like. Enjoy it.

Gut no, data no. Do not proceed. This is clarity, and it is worth celebrating for the time, budget and arguments it saves you.

Gut no, data yes. Pause. This is where bias hides. Ask what is sitting underneath the resistance. A genuine signal, or an old story wearing the costume of one? This is the quadrant where strong-willed leaders who happen to be good storytellers do the most damage, because they can talk themselves, and their boards, out of the evidence.

Gut yes, data no. Caution again, with one exception. If the data says no because there is no data yet, no precedent, no comparable case anywhere, you may be standing in the innovation zone. This is where Airbnb and Netflix were born, and where Opnuu lived. It demands discipline: validate fast, hunt for disconfirming evidence, and understand exactly why your instinct fires before you bet the house on it.

The word that fixed it.

The grid had a flaw I could not see until Marco Iannone, Chief AI Officer at Virgin Active, said a single word on a call as we revised the model together. Good.

Not “what does the data say” but what does the good data say. The version that is current, complete, and actually measuring the thing you think it is measuring. Half my “data says yes” readings had been resting on data that was none of those. A confident number from a stale or partial source is not a green light. That one qualifier changed how I use the grid more than anything else.

The map I found already drawn.

I built this grid from my own scar tissue, so I was startled to later discover a similar four-quadrant structure in Eugene Sadler-Smith’s Trust Your Gut, a traffic-light model he attributes to the Care Quality Commission’s thinking on the role of intuition in professional judgement. This time a traffic light sistem. Almost the same four verdicts. Red, avoid. Amber, check your data. Amber, stress test your intuition. Green, proceed.

A healthcare regulator and a tech CEO arriving independently at the same map is the strongest sign I have that the structure is sound rather than personal. And tellingly, their 'check your data' amber says exactly what Marco's one word says. When the gut resists what the numbers claim, interrogate the numbers before you trust them. If the numbers hold up, interrogate your biases.

Where AI changes the grid.

The grid assumes you have data worth weighing. Often, as I did with Opnuu in 2010, you do not, or you cannot gather it at the scale and speed a decision needs. This is where assisted intelligence earns its place. Never to make the call, but to fill the data axis honestly, fairly completely, and with enough rigour to be genuinely useful. It can pressure-test a market-readiness assumption in an afternoon that would once have needed a research budget I did not have. It can hunt for the disconfirming evidence your own confirmation bias is quietly steering you around.

Used well, it strengthens the very axis Marco told me to improve.

What this means for you this week.

Take one decision you are sitting with now and place it on the grid. What does your intuition say, what does the good data say, and which of the four squares are you standing in?

If you are in a green or a red, act, and trust the time it saves you. If you are in an amber, the grid has done its job. It has told you precisely where the remaining work is: interrogate the data, or interrogate you. Know which question you still owe yourself before you decide.

What is coming next.

Over the next couple of weeks, a few short, science-backed pieces on how to create the optimal conditions for your intuitive signals to make themselves heard. We cannot switch intuition on at will, but there are small, practical things that make the quiet signals more likely to show up.

And if you would like to say hello in person, I am chairing the AI and Innovation stage at MAD//Fest London from 7 to 9 July at the Old Truman Brewery. Come and find me and stay tunes for a miniseries on intuition, data and AI working together I will write straight after the event.

So here is the deal. Your intuition will not come when called, but you can lay out everything it likes, the rest, the walk, the quiet, and wait. And while you wait, check your data, so that when the signal does turn up, it is meeting something definitely good.

If you found this useful, share it with someone about to bet the house on a hunch, or the one who insists the spreadsheet is always right, two ends of the same spectrum, both of whom could do with listening to the input they keep ignoring.

If you are not yet following Intuition Edge, you can do that here:

Subscribe now

Share Intuition Edge

substack.com/@intuitionedge

Published every Monday.

I have always been rather good at avoiding regret, in the sense that I could not name a single one until I was thirty-three or so. Then the big one came. A personal decision, made in service of work, that still stings almost fifteen years later.

The Almost Royal Wedding

It was a wedding in Sudan that I missed. Shaza, a great friend I met in London, and a princess of the Nuba Royal Family in Sudan, was marrying the wonderful Declan over a five-day royal wedding. I said no, I can’t, because I was at home, hunched over a start-up I had convinced myself needed me fully.

I told myself I was backing my gut feeling by doubling down on my start-up. Some of it was, yet the larger part was confirmation bias camouflaged as intuition. Someone who knew the market well had handed me a clear warning, and rather than test it, I reframed it as narrow thinking so I could conveniently keep believing what I already believed. The signal and the self-justification felt identical from the inside. I had no way to test his warning against wider evidence, and in truth I did not much want to.

The Long Pause With Dyson

Last year, as I was running a keynote with the Dyson team, someone put THE question to me directly, for the first time in a live session: “How do you tell a genuine intuitive signal from a conditioned one?” My honest first answer was that you cannot cleanly separate them, because a large part of intuition simply is conditioning, patterns laid down by experience until they surface without conscious thought. A discouraged pause spread across the virtual room, as though we had bumped into a real dead end. If intuition is just conditioning, how can any of it be trusted?

But it can be. What matters is whether that conditioning was formed under honest feedback, or in the comfortable conditions that let you delude yourself.

My friendship with Shaza and Declan had years of evidence behind it, and showing up for someone I love has never once steered me wrong. That was high-quality conditioning, and I overrode it for a venture that had earned no such trust. Luckily for me they understood, and the friendship outlasted the start-up.

I have since built a framework to weigh intuitive signals and analysis together, grounded in what these two scientists agreed on. Miss either condition, and the most convincing signal is just confident noise. Get both, and your sensing has earned its seat at the table. Here is how to tell the difference.

The Two Intellectual Opponents Who Couldn’t Disagree

In 2009, after years arguing from opposite camps, Gary Klein and Daniel Kahneman sat down to settle it. Klein had built his career studying intuition where lives are on the line, making the case for its power while charting its failures as carefully as its successes. Kahneman thought intuition was mostly a trap, a reliable source of errors that can be made all the more dangerous by how confident they can seem. He demonstrated how systematically our judgement misleads us, and his most recent book, Noise, relegates intuition to the same bucket as conditioning, fear and anxiety.

What they published instead was a paper titled “Conditions for Intuitive Expertise: A Failure to Disagree.” It came down to two conditions.

Condition One: Is The World Regular Enough To Be Read?

Intuition is pattern recognition, and a pattern can only be learned if it is actually there. Some environments are regular enough to contain real, repeating structure, from a chessboard obeying fixed rules to a familiar negotiation running along grooves that recur. In worlds like these, enough exposure lets the nervous system absorb the regularities and send you an alert when an anomaly pops up.

Other environments are too noisy, too random, or too novel to hold any stable pattern at all. Klein and Kahneman called the first kind high-validity environments and the second low-validity. The distinction is brutal and clarifying. A firefighter's sensing is trustworthy because fire obeys the laws of physics, whilst a stock-picker calling next quarter's price is not because short-term moves have all the consistency of a toddler deciding what they want for lunch.

The feeling of certainty can be identical in both, but the solidity of the ground beneath it is not, and that is what decides whether it holds.

This is where I came unstuck. A real friendship is about as high-validity as life gets. Decades of the same people showing the same care, an unbroken record of evidence I could read accurately in my sleep. My start-up sat at the opposite end. A market with no direct precedent anywhere in the world, which is precisely a low-validity environment, one with no stable pattern yet formed for anyone to read. My sensing about the friendship was earned, whilst my conviction about the venture was a hunch camouflaged by intuition.

There are exceptions even here: the traders I wrote about who learn to read their own bodies for an edge prove the rule, having built real feedback in one fast, narrow corner of an otherwise unreadable market. Their edge is physiological, not analytical, which is worth holding on to before we hand the hard domains to the machines.

Condition Two: Have You Had Enough Honest Feedback To Learn It?

A regular world is necessary but not enough, as you have to have spent enough time in it, receiving clear and reasonably quick feedback, for the pattern to lodge. What I love about this part of their work is the emphasis on interactions and feedback rather than mere time spent. Twenty years in a role is not twenty years of learning if the feedback never arrived, was never listened to, or arrived so late and so muddied that you could never connect cause to consequence.

Klein’s experts, the fire marshal, generals, nurses, pilots and chess masters, all shared this: they made thousands of reads and found out fast whether each was right. Not every read was. Some failed, a few catastrophically, and in a high-stakes field the cost of a wrong call was lives lost. But that is what made the feedback honest, tight and unforgiving enough to learn from, success and failure both, sometimes through human judgement, sometimes through an aircraft’s black box. Compare that to a domain where the outcome only becomes clear years later, by which time a hundred other factors have muddied the water and your memory has quietly rewritten the story in your favour. You can accumulate decades there and learn almost nothing, while feeling more confident every year, which is the difference between thirty years of experience and one year repeated thirty times. The trap has nothing to do with age, since it catches the grizzled veteran and the excited newcomer alike, and what fails them both is the same thing: a refusal to listen to feedback, which is simply the radical open-mindedness I wrote about last week.

Where AI Changes The Maths

Assisted intelligence can meaningfully help here. It holds more variables than you can, detects faint regularities across data too vast to feel, and surfaces a pattern in the noise where your own sensing has nothing solid to stand on. It can validate a hunch with data, or offer a different reading of it. It is strongest where you are weakest, which is the whole argument for using both rather than choosing between them. Obviously hugely influenced by the questions we ask, the context we give, the taste or judgement we apply.

There is a second, quieter use that matters more. Condition two failed me because back in 2010 the market was too new to produce honest feedback, and I had no way to test its readiness at scale without a budget a start-up never has. Now, assisted intelligence can partially close both gaps. Hand it your reasoning and ask, genuinely, where the evidence says you are wrong. It will not always be right, and it carries its own biases, trained on the language you feed it, which is why I would always pair its feedback with that of a few well-chosen people.

Most models, left untrained, use flattery and confidence to confirm what you already believe, becoming the most sophisticated yes-man and cheerleader you have ever had. Trained to disconfirm, they become the closest thing to an honest black box for your own judgement. Claude Fable 5, one of the more advanced models I tried before its access was suspended, set up to be a contrarian, took even this to an extreme, psychoanalysing me in ways that were not right nor useful.

So take it all with a pinch of salt and stay in control, cognitively and emotionally. AI is, like every such system, designed to keep you coming back, and that alone should carry its own ‘use with caution’ sign.

What This Means For You This Week

Pick one signal you are trusting and one you are overriding. Run each through the two questions: is its source regular enough to hold a real pattern, and have you had honest feedback there long enough to read it? Are the data good enough and complete enough?

My friendship passed both tests. My start-up failed both, and a small voice knew it all along, which is why I worked so hard to create my framework.

What Is Coming Next

Next week, Red Flags, Green Lights 2.0. The decision-making tool I built after the start-up, the one word a conversation with the brilliant Marco Iannone, Chief AI Officer at Virgin Active, added to it, and why a healthcare regulator independently arrived at something strikingly similar. The 1.0 of this thinking appeared in a piece MAD//Fest published late last year. What is coming is the version with everything I have learned since.

And if you want to find me in person, I will be chairing the AI stage at MAD//Fest London again this year, 7 to 9 July at the Old Truman Brewery. I would love to say hi.

If you are not yet following Intuition Edge, you can do that here: substack.com/@intuitionedge

Published every Monday.

Share

I once held a confident, well-reasoned, morally comfortable opinion. It lasted about five hours. Something told me to go down into the details, and I started sifting through over a thousand voices.

The day I changed my mind in five hours.

A few years ago, I was advising a global social platform on a campaign that featured drag content aimed at teenage audiences. The work was well-intentioned. When the campaign launched, the comments under the promotional content filled up fast. Some good, some bad, a fair number of complaints, mostly from parents. The platform’s first move was to delete almost half of the comments entirely, on the grounds of homophobia. My first move was to agree with the decision.

Those comments were the thousand voices. The standard reading at the time was straightforward: homophobic backlash, bigots crawling out from under their rocks the moment LGBTQ content reached a mainstream audience.

Then I actually read all of them. Some were homophobic, but in my view more than half were not. Parents raising age-appropriateness concerns, as I would probably have raised them myself about any content I did not feel my tween daughter was ready for. The parents’ deleted comments broadly made the same argument: the content should not be served to their thirteen-year-old without a conversation they wanted to have first.

Conflating the two was easier than separating them, and the easier reading would have let the platform off the hook for a real editorial question about who the content was actually for.

I radically changed my position. My view on whether the campaign should exist never moved, yet I backed the parents, who I felt deserved engagement rather than dismissal. The platform held its line. I had lost my battle yet it was not my war to fight.

What I kept was the trade. I gave up a comfortable opinion and gained a better one, slightly against my will.

What the science says.

In 1999, David Dunning and Justin Kruger showed that competence and confidence run in opposite directions: the less able overestimate themselves, the more able underestimate how unusual their ability is. Does that sound familiar? The mechanism is the same one Socrates was pointing at. Expertise is what lets you see how much you do not yet know.

Karl Popper put the broader argument differently. The strongest theories are the ones that can in principle be proven wrong. Science progresses through the breakdown of what it currently believes, not through accumulation of what it already accepts. To know more is to be more willing to be wrong.

I am not suggesting that you should consider any view at all, which would be intellectual paralysis, nor the agreement that all views are equally valid, which would be intellectual cowardice. And you are most certainly allowed to turn selective hearing on when the satnav insists, with total confidence, on a route you know is wrong.

Yet what I am proposing is to foster the discipline of holding your own positions as hypotheses, not conclusions. Including the positions you are most certain of. Especially those.

The harder skill nobody mentions.

How do you hold two positions you cannot reconcile without retreating into “everyone has their view,” a stance that sounds open-minded but is actually fence-sitting? Dante had a place for it: the Vestibule, where the souls who refused to take any side are chased forever after a blank banner, stung by wasps, not quite admitted to Hell because they never committed to anything in life either.

To avoid such a dreadful fate, the solution is simply sitting in the discomfort long enough rather than collapsing the positions into one: staying in the tension longer than feels comfortable, holding both as live possibilities, and waiting before you decide which one you stand behind.

I was confidently drawn to use the image at the top of this piece, a picture I took in Puglia, as a sail is only useful while it stays open to wind it did not choose and cannot control. Sheet it in too hard, lock it to the course you have already set, and the boat stalls. Good sailors adjust to forces that keep changing without ever losing direction, not far from the rock climber in my earlier piece, “The Filter Nobody Told You About.” Open is not drifting, and it is not letting the weather decide where you are going.

The second skill compounds the first. In a room, it means giving space to the voices that are not the loudest. The most confidently argued position is usually the least examined one, and the tentative, half-formed view from the quiet end of the table is often the one that has been thinking about the question longest.

A very senior and accomplished CSO once put this to me in a way I have never forgotten. “Liv, we make a great team. You start from a fairly conservative view, yet you are incredibly open-minded. I start from such a liberal stance, and because of it, I am much more closed-minded.” He was right, and the reason we worked is the point. The practice has nothing to do with where you start politically. Some of the most closed-minded people I have worked with hold liberal views, and some of the most open-minded hold conservative ones. The greatest value was that each of us was the stress test the other could not run alone.

Where it has limits.

The practice does not always resolve into action though. Years ago I worked on a strategy with a brand operating in markets where homosexuality is criminalised. A colleague, who was gay, argued the brand had a duty to push against the culture rather than reflect it by omission. I took his argument seriously and pushed the client hard to take what I thought was the right stance. They were clear it was not their fight to lead. I held both views for longer than was comfortable. In the end, I had to drop the argument. Years on, I am still not certain who was right. Sometimes open-mindedness resolves into a clear decision. Sometimes it resolves into uncertainty you carry for years, and that is part of the practice too.

Open. You can test your baseline here.

Open-mindedness is measurable, and here is the test I promised you. The cognitive scientists Keith Stanovich and Maggie Toplak have spent two decades developing what they call the Actively Open-Minded Thinking scale, and their central finding is worth restating: the trait is separable from raw intelligence and predicts calibrated judgement. Smart people who score low are confidently wrong more often than slightly less smart people who score high.

Here is the thirteen-item version, reformatted from their research. Score each statement from 1 (disagree strongly) to 6 (agree strongly). There is no neutral point, by design. You have to commit, which is rather the joke.

1. People should always take into consideration evidence that goes against their opinions.

2. Changing your mind is a sign of weakness. (R)

3. I like to think that my actions are motivated by sound reasons.

4. It is important to stick to your opinions even when evidence is brought to bear against them. (R)

5. Intuition is the best guide in making decisions. (R)

6. Considering too many different opinions often leads to muddled thinking. (R)

7. One should disregard evidence that conflicts with your current opinions. (R)

8. Coming to decisions quickly is a sign of wisdom. (R)

9. Allowing oneself to be convinced by a solid opposing argument is a sign of good character.

10. If something I think feels right then I am comfortable, whether or not it is true. (R)

11. A person should always consider new information.

12. People should revise their conclusions in response to relevant new information.

13. Certain opinions are just too important to abandon no matter how good a case can be made against them. (R)

For the eight items marked (R), flip your score: a 1 becomes a 6, a 2 a 5, and so on. Add it all up. Scores run from 13 to 78. Higher means more actively open-minded.

How AI comes in handy, and what it told me about working with a friend.

I ran the test for fun with a friend, Bea, a Chief Legal Officer in a sustainable energy scale-up. The results were more useful, and more entertaining, than I expected. The headline surprise was this: disposition does not predict the score. Some of the people I would lazily have filed as open-minded by lifestyle scored lowest. Some of the quietest, most cautious in the group scored highest. The label and the capability are not the same thing, which reminds us again never to judge a book by its cover.

Bea and I diverge in a way that turns out to be complementary. We share the foundations, a strong instinct to revise our beliefs when the evidence asks, and neither of us clings to a position out of ego. Where we differ is what we do with complexity. I expand, most comfortable holding several competing perspectives open at once and pushing toward the truth even when it is uncomfortable. Bea resolves, distilling complexity into clarity and moving toward a decision. On tolerance for ambiguity the gap is widest: I sit at the open end of the scale, Bea at the decisive one. Both incredibly valuable skills, and even better when combined.

Then I did something I would not have thought to do two years ago. I gave both our anonymised profiles to AI and asked a simple question: should these two people work together, and how. The brief was sharper than either of us expected, though it tempted one lazy conclusion I want to correct, because it matters for how you read your own result. A high tolerance for ambiguity is not the same as being slow to act. Anyone who has worked with me knows I am known for the opposite, for pragmatic progress and for sensing exactly when enough deliberation has happened. What the scale actually measures is how comfortable you are holding competing ideas open while you think, not how fast you move once you have decided. On that, Bea and I differ usefully. I am built to widen a problem and pressure-test it. She is built to take a framed decision and drive it into a clean action plan. We are both glad to hand the other the part we each enjoy least.

This is AI doing what this series has argued it does best: it read two patterns we already embodied and handed them back sharper than we could articulate ourselves. The test surfaced the data; AI helped us see its shape.

What this means for you this week.

If you take one thing from this piece, make it a habit rather than an idea. Five, actually, since this is a piece about not settling too early.

1. Hold your strongest positions as hypotheses, not conclusions. Especially the ones you are most certain of.

2. When two views you cannot reconcile sit in front of you, resist the urge to collapse them. Stay in the discomfort and keep testing both against the evidence. The Vestibule is for the ones who refuse to choose at all, not the ones who take their time.

3. In any room, give the quietest voice the same airtime as the loudest. It is often the one that has been thinking longest.

4. Find the person wired least like you and treat the difference as an asset. Set the handover early, so neither of you mistakes the other’s style for a flaw.

5. Take the test. Properly, with no neutral fence to sit on. Then take it again in six months and see what moved.

One provocation before you go. You’ll have noticed item five: “Intuition is the best guide in making decisions,” and that agreeing with it lowers your score. In a publication called Intuition Edge, that should raise an eyebrow. The researchers treat leaning on intuition as a mark of weaker thinking. I think that is both right and badly incomplete, and I am not the only one. Two of the most serious minds in decision science spent years fighting over exactly this question, a sceptic and a believer, and the conclusion they reached together surprised them both. That is where we go next week.

What is coming next.

Next week, two scientists and a meditator. A sceptic and a believer who could not have been more opposed, who sat down for years to settle whether trained intuition can be trusted, and published their answer under a title that admits they failed to disagree. Plus the part the science has not worked out yet.

If this piece resonates, share it with the person you argue best with.

If you are not yet following Intuition Edge, you can do that here: substack.com/@intuitionedge

Published every Monday.

Share

The best work of your career almost certainly did not belong to you alone as no idea is truly original. If you have spent any time in a creative job, you will know it is almost impossible to attribute great ideas to one single human in almost any case worth examining, even if some individuals would love to claim otherwise, and even when that individual is you.

The trick is that the solution is often attributed to the person who named the unlock, cracked the code, or is credited with better taste. Let us start with the latter, which is arguably the most subjective of the three.

My answer to what taste is, is simple yet carries deep implications and subtleties. A person’s taste is shaped by what they have lived, what they see now, and what they want next. Three layers, all of them porous, none of them sealed off from the world. This is somewhat why ideas cannot possibly be original, or attributable to a single person. When an idea is ready to be born, multiple attuned people pick up the signals in media, society, and conversation, each choosing their own unique path to bring it to life.

And nothing crafts a spark into form better than a diverse, well-curated group bouncing off each other.

A Random Collective?

Now, it is not diversity for the sake of diversity. You would not invite a skateboarder into a brainstorm aimed at working out whether the Higgs boson can explain the matter-antimatter asymmetry in the universe. Unless the skateboarder is also a physicist, of course. Diversity inside a base of shared competence is the magic. A shared language to have a debate. An instrument you can play to join the band. Outside it, it is just noise.

So what makes great ideas and great problem solving flourish? Some point to the room’s composition, who you actually get in there. Some point to how the group engages. Some even point to the room itself, the conditions under which it operates. In an attempt to keep this short, I will focus on the first two.

The fact that none of us is as good as all of us has been quietly understood and endorsed by some of the most successful operators and creators of this century and the last.

Ray Kroc used it over and over to define the very structural fabric of his franchise philosophy, McDonald’s.

Not knowing this at the time, I assumed it was an original quote from one of the founders of BBH when I first walked into their fancy London office. It is the kind of insight that turns up in multiple rooms, finds different people ready to pick it up, and gets attributed in good faith by each of them to the person who said it most recently. Which is rather the point.

A small recent example from my own work. I wrote a piece a few weeks ago suggesting that rock climbers make great problem solvers, and I genuinely believed the idea was mine. After publishing, I shared the piece with Chris, a neuroscientist friend whose feedback I trust, and he mentioned that surely I had been inspired by Flow, by Mihaly Csikszentmihalyi, where rock climbing is used as a central example. I had not actually been inspired by it. But it made me remember that what sparked the idea was a conversation with a very talented head of project management I worked with, who was a passionate rock climber. I had completely forgotten about that conversation until Chris’s reference forced me to trace the thought backwards. While I was writing the article, it had felt like my original idea and conclusion. It was not. Ideas float, get picked up, discussed, made better by the collective.

But not by any random collective.

Higher IQ Does Not Make Best Contributors

Being socially intelligent does.

Anita Woolley and her colleagues at MIT and Carnegie Mellon studied what makes some groups consistently outperform others across a wide range of tasks. Their finding was counterintuitive. Group performance was not strongly predicted by the average IQ of members, nor by the IQ of the smartest person in the room. What did predict it was the average social sensitivity of the group, measured through a test of how accurately people read each other from minimal cues. Groups whose members were better at reading each other outperformed groups of higher individual IQ. Equality of conversational turn-taking mattered too, as did the proportion of women in the group, partly because women on average scored higher on social sensitivity. I wanted to be surprised there, in favour of all the sensible and socially attuned men I know, but I would be a hypocrite to say I was.

What is even more interesting is that the effect held even when teams were working remotely rather than face to face. The room reads itself. The better it reads itself, the better it performs.

Scott Page at Michigan came at the same question from a different angle and showed, mathematically, that the collective accuracy of a group improves with cognitive diversity even when individual accuracy is held constant. People bringing different mental models to a problem search different parts of the solution space, so the group covers ground no single mind could. Crucially, this only works inside a shared base of relevant competence. Outside that, diversity is noise rather than signal.

Matthew Syed, makes the contemporary case that cognitive diversity is the most under-leveraged asset in modern organisations. Most teams suffer not from a lack of intelligence but from perspective blindness, the fact that we are oblivious to our own blind spots. A team made of people who think the same way will share the same blind spots and, I would add, carry the organisation’s biases on top of their own individual ones as members of the organisation. A team made of people who think differently, inside and outside the organisation yet all within a shared frame of competence, will catch what each member alone cannot.

Three findings, three angles, one underlying argument. The room is the unit of intelligence. The individuals are inputs. Messi’s input to Argentina is a great one. It is still just one input.

And this is where everything else this series has built starts to compound. The pattern recognition, the somatic signals, the atmospheric sensing, the saliency filters, the heart and gut and body reading what the conscious mind cannot. All of it. Individual intuition does not just survive the move from the solo mind to the collective room. It amplifies inside it. The signals get sharper because they are being read and refined by multiple nervous systems in parallel. The room hears more than any one person in it can.

The Career Hack Nobody Told Me About

In 2003 I left an offer with the Publicis group, dropped from director to senior manager, and took a 25 percent pay cut. The reason at the time felt simple. I was watching incredibly smart marketing directors around me freeze in front of data, and I did not want to be one of them. I wanted digital under my belt before my career went any further.

I joined Tribal DDB as a senior account manager. Which was a dent to my ego as I went from a Director role back to a Senior Manager one and a salary reduction of about 20% that made up for the fact that I wasn’t a terribly good one on top of it. Coming from the commercial events and partnerships space, the learning curve was steep as I did not even know the difference between HTML and a .gif. But I could think pragmatically about strategy, was a good listener and had a can do attitude, so I was put to work as a liaison between DDB and Tribal, spotting opportunities for briefs to be carried across from the advertising agency downstairs to the digital agency on the third floor in a way that made sense for both. I was running a lot in stiletto heels (yas that was back in the days where trainers were not quite cool in my eyes) up and down between floors a lot.

Jokes aside, that route gave me the chance to work on one of the most complex, multifaceted global brand repositionings I have ever been part of. Philips moving from “inspired technology” and being an engineering led company, to “Sense and Simplicity” and being more customer led, across every category, with the change made applicable and brought to life across every channel, including the ones I was still learning.

Not even two years later I was back to director, and a group one on top, on double my original salary, with a global remit. That was the surface gain. The deeper gain took me fifteen years to fully understand.

What I had actually learned in digital was not how to talk about HTML. It was how to bring genuinely diverse talent together. Designers, software engineers, data specialists, UX and UI, copywriters, product managers, the whole spread. Digital forces this in a way that branding, events, partnerships, advertising, and pure strategy work do not. You can produce excellent work in those sectors with teams that are competent and culturally close. You cannot produce a category-defining digital product without orchestrating people who think, work, and speak in different professional languages.

The hack was not the skill I thought I was buying. It was the diversity I was forced into. In my view it gave me an essential skill for any good CEO. Learn how to curate a great team and give them the space to collaborate and co-create the best work of their life.

If you are mid-career and looking for the next move, this is the version of the advice I wish someone had given me. Do not deepen the muscle you already have. Move into a context that forces you to work across more cognitive diversity than you currently do. The new skills are useful. The ability of curating diversity can become your edge.

Where AI Fits In The Room

AI is the newest participant at the table. Used well, it adds another voice with different blind spots, which increases the cognitive diversity of the group. Used badly, it replaces the room and produces fluent, confident, single-perspective output that no one in the group challenges because it arrived with the authority of the LLMs that is designed to sound confident and tell you what you want to hear.

A short example. I have spent time with the team at Virgin Experience Days on how to integrate AI into their decision making without losing what makes their team genuinely good at reading the market. Danni Rush, the CEO, was clear from the start. AI is not the oracle. The team is.

AI is invited into the room as a perspective the team can interrogate. It is also an enabler, supporting data modelling and research coverage that would otherwise be too inefficient to produce. In practice, this means using AI to surface what the team might be missing, to play back counter-arguments to a position they are already leaning into, and to offer angles from outside the room’s frame. The individuals challenge both the room and the AI to sense what neither can sense when the work calls for genuine novelty.

This is the cleanest framing I have for where AI belongs in collective sense making. It is another seat at the table. Never to be mistaken for the table itself. The team’s social sensitivity, turn-taking discipline, and cognitive diversity drive and decide.

The Practical Bit

Our ego plays tricks on us. It makes us believe we came up with genuinely original ideas, or that our contribution to a brainstorm was greater than the input from the collective. Both are almost always wrong. The honest move is to brainstorm with a curated set of people, practice radical open-mindedness (which I will write about more another time), and accept that the outcome belongs to the room.

A jazz band works off each other’s cues for brilliant improvisation. Argentina’s national team moves as one, even with Messi on the pitch. Your ideas get better when perspective is actively sought from the right group, and worse when you sit alone with them trying to think your way out.

This is what I curate and facilitate in the workshops I now run with clients. It is what makes the work better. If what you are after is long-lasting, better decision making in service of better outcomes, focus on curating the right teams and creating the right space for them to succeed in collaboration and co-creation.

Restrain the ego. The temptation to behave like a lone, fierce, majestic tiger is real, but it will not serve you well. Harness the power of more humble yet strong ants. Your ego might not like it yet your outcomes will.

What Is Coming Next

Next week we move from the room to what the individual does to participate in it well. Radical open-mindedness, why it is harder than it sounds, and how to practice it without becoming the most insufferable person in the meeting.

In the meantime, if you are curious about where your own intuitive practice sits right now, I have built a short self-assessment that takes a few minutes and gives you a starting point. Get in touch with me if you want to have a play with it.

If this piece resonates, share it with someone who has ever taken too much credit for a brainstorm. Or, even more usefully, with someone who has quietly carried a brilliant room and never been told.

Published every Monday.

Share Intuition Edge

“You’re on your way to a heart attack by age 50,” he said, when I was 39. “Or you might live to 100, if you take five minutes every day to do what you just did here today.” These were the words of a highly experienced Scottish doctor, at one of those two-day executive health checks my old company sent me for, the kind that measures everything, a scheme belonging to a time not too long ago when those things existed for employees.

Here I was, a woman in her late thirties, living sensibly, healthy in every measurable way, and yet my stress markers were off the chart. He was visibly cross at the colours from the readout in front of him, offered by a machine measuring my heart rate variability. The chart showed far too much red, a lot of blue, and just a little green.

I had no clue you could actually see stress back then, and I was fascinated, thinking all the while that today was a good day on top. My body had been keeping a record I knew nothing about, now laid out in colour for me to see on a screen.

This is the same physiology I wrote about a few weeks ago in the heart piece. Back then it was John Coates talking about heart rate variability and how this affected the best traders on the floor. Here it was me, a chart, and a slightly scolding Scotsman.

He asked me to retake the test. This time he told me to breathe in for five, out for five. Close my eyes and picture something nice. Then he left the room and said he would be back in five.

I did as I was told. I breathed. I closed my eyes. And after maybe a minute, my mind went somewhere I did not expect and have never quite been able to explain. A dark space, scattered with points of light, waves and tiny sounds, where the only task was to resist the pull to decide there was an up or a down, a left or a right, and instead to stay merged and dispersed in all of it. When I later described this to my husband he was quite disbelieving and fairly annoyed, on the grounds that I had reached this state without the assistance of mushrooms, which he felt was somehow unfair.

The doctor came back and was startled. The chart was now around 95 percent green, with a thread of blue. A completely different picture from the one he had left. He told me, with a clinician’s flourish rather than a statistician’s precision, that he had rarely seen anyone move that far that fast, and put me somewhere in the top couple of percent he had ever tested. To live to a hundred, he said, I simply had to sit with myself for five minutes a day and recreate what I had just done.

I want to pause here, because I am aware of how this reads. It is genuinely difficult to write about scoring 95 percent green and being told you are built to last without the entire room quietly hoping you trip on the way out. We do not warm to the person who turns out to be top of the class at the one thing they were not even trying at. I thought about softening it, or leaving it out. But the whole point of this piece lives in what happened next, and the relatable part was never the gift. We all have different ones. It was that I had this specific gift, and the proof, and the prescription, and I still could not do the one thing I had just been told would change my life. Even with two young daughters and a husband I deeply love, and every reason to want to be healthy, over and above my own deep love of life - even when it is less than perfect.

So I am telling it straight. Choosing honesty over likability here. Because what came next is not really a story about meditation. It is a story about how the answers we cannot reach finally arrive.

Meditation Attempts, Pathetic At Best

I signed up for meditation classes. I hated every second of them. I did not keep it up, and not just for a few weeks but for four years.

It did not matter how much I understood the science. It did not matter how many times I picked up James Clear’s Atomic Habits, or how many expert friends I asked for help. It did not work, because I did not do it.

When I knew I was about to leave my job, I arranged the same full health check again to make the most of a benefit I almost forgot I had. Same company, different doctor. This time a practitioner in her mid-twenties. She found me slightly dehydrated and otherwise much as before. Excellent across the board, stress markers alarming.

She asked me to retake the stress test, breathing five in and five out for five minutes. But she did not ask me to close my eyes. She did not tell me to picture anything. And she stayed in the room. The result improved, by perhaps ten or fifteen percent, but nothing like the transformation of years before.

I got curious. I asked whether she would let me do it once more, this time properly alone, eyes closed, for the full five minutes. She agreed and left. When she came back she was visibly shaken. The chart was entirely green. She said she had never seen anything like it. Which, I suspect, says more about how early she was in her career than about me, but the point stands. Given the right conditions, my body did the thing instantly and apparently very well.

So now I knew it twice over. No excuse left. If I wanted to be well, I simply had to find five minutes each day. And for another year after that, I read, enquired, and wasted money on a pay-as-you-go set of meditation classes I did not attend, spending most of my time feeling guilty about my failures. Full awareness, zero change. It is a particular kind of frustration, knowing exactly what would help, feeling lucky to be able to do something you have been told is quite hard, and watching yourself not do it, week after week. It is like knowing you are going to win the lottery for sure and finding excuses not to buy a ticket. I felt like a real idiot.

Why Nothing Was Happening, And Everything Was

Here is the part I did not understand at the time, and only understood later through the science.

As I thought I was going nowhere, I was in fact subconsciously accumulating hints, cues and suggestions that would later help me solve the “why am I not making time to meditate to feel better” problem.

There is a mechanism behind the eureka moment, the answer that arrives unbidden in the shower or on a walk, and it was given its fullest recent description in a 2025 paper in Communications Biology by Steven Kotler, Karl Friston and colleagues. Brace for it, as it is by far my favourite piece of learning this year, even if I had to read the paper three times to fully get it.

Opportunistic Assimilation

When you hold an unresolved problem, your brain does not file it away and stop. It keeps it quietly active in the background while you seemingly forget about it and get on with your life, yet it uses that open question as a filter, scanning everything you subsequently encounter for anything that might resolve it. The problem recruits new information, continuously and below the threshold of awareness, until enough of the right pieces assemble for the answer to surface as our beloved metaphorical light bulb.

The crucial detail is that the answer is rarely new information but rather a new arrangement. The pieces were mostly there already and what was missing was the configuration and, most crucially in my view, the readiness to receive it.

My year of guilt and reading and asking was the background process running. I just could not see it working, because by definition you cannot. The surface is clear. The depth is doing the work.

A Friend, Emotions And A Beach

It resolved on a beach, late in the evening, in June. As we were pouring our hearts out to each other, each with our own particular messes to deal with, my wise friend Alex asked me a simple, direct question:

“Are you sure you want to get better?”

It landed like a punch to the stomach. One of the key somatic signals I have spent this whole series describing, arriving with utmost clarity before I could engage anything else. I knew immediately she was right. I just could not see how, because it made no sense to the pragmatic, optimistic, make-the-most-of-it version of myself I thought I fully was.

I had two daughters I adored, both being affected, especially the youngest one, by my high stress levels due to a horrible exit from a company I was going through. If you are senior enough you have most likely had one of those exits too. As my mentor Peter Slator reassured me at the time, let us stop the LinkedIn veneer that those things never happen, or when they do they cannot ever be mentioned. But what is actually important to this story is that I was not using everything I had to get better and manage this incredibly stressful period of my life. For me, for them, and especially for the youngest one.

Here is the detail that matters most for the mechanism. A couple of months earlier, someone I had only just met that evening, someone I felt comfortable enough to speak openly with, had asked me almost exactly the same question. And it had slid straight off me. I had genuinely forgotten it had been said until this other, newer friend reminded me much later. By June, on the beach, with a year of background processing behind me, I was finally ready to receive it, from someone I trusted as a long-standing friend with whom I had spent a great deal of professional and family time. She was almost my boss for a moment, years before, but that is a story for another time. The information had not changed. I had. The signal needed the right conditions and a mind that had accumulated enough to catch it, which is precisely what opportunistic assimilation does.

The revelation itself did not arrive in the conversation but the next morning, as I woke, which is the second classic marker of this mechanism. The conversation seeded it. Sleep assembled it.

A Bag Full Of Daddy Issues, Unresolved

What did surface? As a teenager I had needed to believe my father loved me, even though he never did the things I thought he could have done to get himself better. For me.

I had been quietly proving that case for years by repeating it, declining to get well myself while loving my own two daughters completely. It was me saying to my thirteen-year-old self, look, your father loved you, you love your daughters and yet you are not getting better either, you see? Loving was not in question, it hardly ever is. The reassurance that thirteen-year-old needed, in facts rather than words, was.

I will leave the full shape of that where it belongs, which is not here. But the moment I saw it, the resistance that four years of evidence and a year of effort had not touched simply dissolved. What surfaced, more clearly than anything else, was a loud “Fuck that!”

I have meditated almost every day since.

Where AI Actually Fits

AI can ask a sharp question, sharp enough to land somatically. This is part of what I train individuals and teams to do, respond to those sharp questions when they arrive. No surprise, then, that its use for emotional support has risen dramatically in 2026.

The slow work that makes you ready to receive the answer is yours and yours alone, as AI cannot hold your problem open for a year while you live your life. It cannot accumulate inside you. It could not have been Alex on the beach. Not because of what she said but because of who she had been to me for years before. The relational and temporal substrate that opportunistic assimilation runs on is not something AI enters. The work was human, embodied, temporal, partly unconscious, and partly relational, and there is no prompt for any of it.

AI did, however, do something else genuinely useful for me. As I train companies and teams on how to use intuition to engage AI better, I always promote meditation as a practice that quietens the noise so you can better hear the signals. I am not an expert in meditation and I often got asked what type of meditation I practiced. I did not know as I had no clue what it was. I wish I could have said 'Transcendental.' Easier to get and hugely fashionable. Instead, I simply went to that dark space full of wavering light and tiny sounds and stayed there, merged with it all.

So I described the experience to ChatGPT, as plainly as I could, and answered a few questions posed by the algorithm.

It told me I appeared to have arrived, on my own, at a hybrid of open awareness meditation, non-dual practice, and formless absorption, with the closest traditional parallels being Dzogchen and Mahamudra, and some overlap with the formless states described in certain Buddhist traditions, particularly those concerned with infinite space and infinite consciousness. It noted, more strikingly, that I seemed unusually sensitive to the constructed nature of spatial orientation and the division between self and other, perceiving their assembly and disassembly directly rather than understanding them as concepts, and that this was relatively uncommon. Please do not ask me again in a keynote or workshop session, as I cannot even pronounce Dzogchen and Mahamudra, but I am glad I now have the answer.

I had reached the destination with no map. The machine handed me the name of the country.

The temporal, embodied, relational work of a human life is not something AI competes with, because it cannot enter it. What AI does, superbly, is meet you on the other side and help you understand what you already found.

What To Do With This

The practical payoff is uncomfortable, so I will state it plainly.

Awareness is not readiness. Knowing what would help you, even knowing it with biometric certainty, does not move you.

You cannot force the eureka moment, nor know for certain what it will deliver when it arrives, but you can build the conditions that invite both. Hold the real question, not the convenient version of it. Allow the background process to deliver. Sleep on it. Have a conversation about something else. And stay genuinely open to the people slightly outside your problem, because the person who asks you the question that unlocks everything is rarely the expert and rarely on the second attempt.

The most important question anyone asks you may be one you are not yet ready to hear. You may need to hear it twice, from two different people, a couple of months apart, before it lands. And that is the mechanism working. Your job is to keep yourself in the rooms and the conversations where the right question might get asked again. Eventually you will have accumulated enough for the bulb to light.

I had the proof for five years before I changed. And the proof was not enough. But a question on a beach was.

Writing this took more out of me than I expected. So just this once, an affectionate nod to three people who made it possible. Alex Dimiziani, for asking the question. Dr. Peter Houghs, for giving me the courage to write this piece, and for handing me my next unresolved problem to sit with. And Rick Rubin, for being my favourite read when meditation and creativity are at play.

What Is Coming Next

Next week we move from the individual to the collective. Why a group can sense things no single person in it can, why genuinely diverse teams read a situation more accurately than brilliant individuals working alone, and where AI fits as the newest participant in collective sense making.

If this resonates, share it with someone who knows exactly what would help them and has not managed to do it yet. They are not failing, even if they know Atomic Habits by heart. They may simply not have heard the right question, from the right person, at the right time, just yet.

If you are not yet following Intuition Edge, you can do that here: substack.com/@intuitionedge

Published every Monday.

Subscribe now

Klein’s firemarshal saved his crew by reading a burning building. You do it every day in rooms that are not literally on fire, and you probably do not give yourself nearly enough credit for it. Every meeting you have ever walked into, your nervous system was already running its analysis before you took your seat, reading micro expressions, energy, the quality of attention people were bringing to their chairs, the texture of what was unresolved in the air. You were sensing the room. And the room, without exception, was sensing you back. Even when your mum is not in that room.

This article is about both directions. About what you pick up before anyone speaks, what you broadcast without knowing it, and how the best leaders learn to work deliberately with the full signal rather than just the half they can consciously control.

What Your Body Registers Before Your Mind Arrives

In the mid 1990s, a psychologist named Nalini Ambady at Harvard ran a series of experiments that became known as thin slicing, the study of how accurately people make judgements from very brief, sometimes silent, exposure to another person’s behaviour. In one study, participants watched three ten second clips of a teacher with the sound turned off and rated them on a series of qualities. Those ratings correlated at 0.76 with the end of semester evaluations given by students who had spent an entire term in that teacher’s classroom. Three silent clips of ten seconds versus one entire semester of lived experience. Yes, the conclusions were almost identical.

Pattern recognition operates at extraordinary speed, reading micro expressions, posture, energy, the quality of attention a person brings into a room, and delivering its assessment before conscious thought should be authorised to form a sentence about any of it. The body registers the full human being in front of it, not the version they are presenting but the version they actually are, and it does so almost instantly.

What this means in practice is that you already know more about the people in any room than you think you do. The real question is, are you paying attention to what you already know?

The Meeting With The Ad Agency’s Client

There is a particular kind of meeting that anyone who has ever been part-handed a client or referred into a relationship by a parent company will recognise immediately.

The client believed they could spend very little, having been advised as much by the leading party who held the primary relationship and had every interest in retaining the lion’s share of the budget. They were also not quite sure what they needed. The leading partner was keen to ensure the budget stayed within their orbit. And a specialist intermediary consultancy, engaged specifically to help the client make the right decision about where their budget should go, was doing its best to hold the process together. Everyone was trying. Nothing was moving.

I was working at ZAG, a spin-off of BBH, and one of my partners had been managing every layer of this particular situation with considerable skill and considerable exhaustion. I asked to be invited to what felt like it might be the final attempt before all parties agreed, politely, to stop trying. He graciously welcomed me into what was, by any measure, a very delicate room.

I walked in and within moments I had a strong read of what was happening. A particular unease from the client, not hostility but the wariness of someone who has been through too many meetings without gaining enough clarity to grant progress.

Tension from the intermediary consultants, doing their professional best to hold something together that kept slipping. My partner exhausted in the way that only comes from months of trying. The BBH partner half annoyed, half quietly relieved, because if the budget stayed small it would most likely stay with them anyway.

And then I noticed the woman sitting opposite me to my left. A small exchanged glance, nothing more than a fraction of a second of genuine attention, gave me the feeling that she was reading the room the same way I was. I made a decision to seek her out as an ally.

I started asking questions. She asked some that were complementary to mine, opening space for me to expand, and as she did I was also gathering more precise information that was evolving my understanding of the situation in real time. The more I understood, the more specifically I could speak to what the client actually needed rather than what the agenda assumed they needed. I left that room with a client who was prepared to be convinced to spend five times what they had originally envisioned.

Four weeks later, after many rational decks and a great deal of follow through, I had the account. Two years later I had a genuine friend in Claudia, the woman who had been sitting opposite me to my left, with whom I went on to work and collaborate many times.

This is what Klein’s firefighter and the art historians at the Getty have in common with every leader who has ever walked into a difficult room and known immediately what was needed. The mechanism is the same. The application is different. And it is entirely learnable.

The Room Reads You Too

“Liv, you are like the weather. When your mood is good you energise everyone around you, like the sun. But if you are tense, frustrated or impatient, we feel you so much, like the worst of storms is about to unleash on us, even if you do not say or do anything.”

My team told me I affected them like the weather, even without saying or doing anything.

And this is the part most writing on atmospheric sensing leaves out entirely: the room had been reading me the entire time. Every meeting I walked into carrying unresolved tension. Every moment of impatience I thought I was containing. My team, my peers, my board had been registering all of it, not through anything I said but through exactly the mechanism Ambady's research describes. From micro expressions to posture to the quality of attention I brought into the room, or failed to bring.

It is quite obvious once you say it out loud. Of course sensing is never one directional and you are always being read, whether you are paying attention or not. This is not to make you too self conscious or careful to mask your emotions at all times, we are human and vulnerability is a sign of strength not weakness. Yet trying to bring your best energy to the fore, or perhaps having the self awareness to say “today is a bit of an off day” might allow others to lean into what they are sensing with a context that is more measured and constructive.

Reading the room is important. Allowing yourself to be read with context could become a great asset.

People Buy Aeople and Teams Read Teams

When Neil presented Make Waves for the Yamaha pitch in the BBH conference room, with the Japanese and Austraian clients sitting across from us and everything riding on the next few minutes, what the clients responded to was not only the idea. It was the relationship between Neil and me. The quality of attention we paid each other. The shorthand of two people who had worked together well for long enough. The ease and the trust and the creative generosity between us that was visible in the room before either of us had said a word about the work.

One of the clients told us afterwards that what had impressed them as much as the campaign was how Neil and I interacted. They wanted people who genuinely got along and complemented each other the way we seemed to. They were buying the relationship as much as the idea.

This is a version of atmospheric sensing that operates at the collective level, and it is more powerful than anything an individual can generate alone. It was operating in both directions simultaneously, they were reading us as we were reading them. And the signal that passed between all of us in that room, before the rational evaluation of the work had even properly begun, was what made everything else possible.

Teams are stronger than any individual. Not just in terms of capability but in terms of the signal they generate together. A team that trusts each other, that has developed a genuine quality of collective attention, reads and is read completely differently from a collection of talented individuals operating in parallel. The clients felt it as the best clients often do.

Sensing Versus Hyper-Vigilance

There is an important distinction worth making before we go further, because the two can feel similar from the inside while producing completely different outcomes.

Trained atmospheric sensing is curiosity-based. It arrives as openness, a genuine interest in what is actually happening in the room rather than what you expected or feared would be happening. This is the key distinction between intuitive and conditioned signals that many people ask me about in our workshops, and one I have developed a specific framework around to help people tell the difference in real time. It tends to produce connection, insight, and the kind of flexible response that a situation actually requires rather than the one your anxiety, conditioning, or biases prepared you for.

Hyper-vigilance is fear-based. It arrives as threat detection, a nervous system running permanently on alert, scanning for danger rather than for information. It can be exhausting. It tends to produce defensiveness, misreading, and the kind of over-interpretation that turns a colleague’s neutral expression into evidence of hostility that was never there. People who had a specific type of upbringing, growing up with an alcoholic or mentally ill parent, experiencing verbal or physical abuse, or parentification, having to step up and assume the role of the parent at a young age, are particularly predisposed to hyper-vigilance. And I am well placed to vouch for this from direct experience.

Yet I learned to tell the difference, albeit not always obvious in the moment, and it is a valuable distinction. If reading a surrounding leaves you more energised and more connected to the people in it, you are developing a capability. If it leaves you exhausted and more certain that everyone around you has hidden motives, you may be running a different system entirely, one that deserves some gentle attention rather than further training.

The goal is not to read more but to read more accurately, with less noise and more genuine curiosity about what is actually there.

What AI Cannot Do In This Room

Assisted intelligence can process data about human behaviour at extraordinary scale. It can identify patterns in language, detect sentiment in text, and flag anomalies in communication that human attention might miss. Which we can’t deny, are genuinely useful capabilities.

It cannot feel the shift when an ally locks eyes with you across a table before a word has been spoken. It cannot register the quality of attention a person brings into a room, or the precise texture of tension that has been building for longer than that morning, or the difference between a pause that means someone is thinking and a pause that means something important is about to not be said.

This is not a marginal limitation of current AI that will eventually be solved. It is a description of what embodied human intelligence is. The kind that evolved over hundreds of thousands of years precisely because reading other people accurately, in real time, in complex social situations, was the difference between survival and its absence as Malcom Gladwell tells us in Blink. We all have that capability and it deserves the same deliberate development you would give any other professional skill.

What This Means For You This Week

Before your next significant meeting, take two minutes to notice what you are bringing into the room. Not what you are planning to say, but what you are carrying, emotionally, physically, in terms of the quality of attention you are about to offer the people who will be reading you before you have opened your mouth.

Then, during the meeting, notice one thing that is not being said. One genuine observation about what the room is actually doing rather than what the agenda says it should be doing. Write it down afterwards and over time, these observations will become one of the most reliable sources of intelligence you have.

The room has always been reading you. You might as well read it back.

What Is Coming Next

Next week we explore the science behind your best ideas and why they almost never arrive at the desk. Unresolved problems recruiting new information subconsciously. The conditions that either invite insight or prevent it. And why the eureka moment is not random at all.

If this resonates, share it with someone who walks into rooms and knows things they cannot immediately explain. They can now use this article to examine it and fence off the doubters.

If you are not yet following Intuition Edge, you can do that here: substack.com/@intuitionedge

Subscribe now

Published every Monday.

Climbers make exceptional problem solvers. I believe great problem solvers look at a problem from a distance, plan and yet adjust as new information arrives. Before a climber takes the first move on a new wall, they stand at the base and look up. They mentally simulate the ascent, planning the route from the bottom, tracing a line through the holds, building a mental map of how to get from the ground to the top. It is a good plan yet most often an incomplete one, because the wall, the environment, and your body have information that will only be shared once you are up. Are you asking why I did not use a picture of something climbing? Patience.

As the climb begins, new data arrives continuously and from every direction simultaneously. The wind picks up slightly and the surface becomes marginally less predictable. A hold that looked solid from below feels less certain under actual grip. A patch of shadow reveals moisture on the rock that was invisible from the ground. Perhaps the right hand registers less strength than expected. The body is processing all of this in real time, some of it consciously, much of it below the threshold of awareness, and the plan formed at the bottom is quietly, continuously being revised. So the art of not falling is to move fluidly from conscious to subconscious and adjust your climb as new information becomes available.

This is the best example of saliency detection for me. Not a single moment of noticing but a continuous, dynamic practice of staying open to what the situation is actually offering, updating the model rather than executing a fixed script, that might have even been well informed from the outset yet is inevitably incomplete.

The best problem solvers never stop reading the wall

They plan from the available information and remain genuinely receptive to what arrives next.

I have been thinking about this a great deal lately, partly because of a project that taught me the cost of losing it, and partly because of this photograph I nearly did not take.

At the end of last year, I took over the overall management of a project for Go-Ahead London, one of the largest bus operators in the capital, at a moment when the release of a suite of real time operational tools was stalling, despite being run by a talented, seasoned team that had been working hard. There was nothing wrong in plain sight as the bugs were being fixed and the reporting was in a decent shape, and yet the CEO’s confidence was not building as the teams on the ground were not rallying. No surprise that the sentiment around the release was not shifting in the way it needed to.

I let the team run with their approach for a few weeks, watching carefully, trying to locate the source of the discomfort. The work I was receiving back was correct. It was also, I kept feeling, not quite right. The tension this created was palpable. My head of product management was giving me their best, from their perspective, and I kept pausing on it without being able to explain why. I was also working with an incredibly talented Programme Manager I had worked alongside in different companies for over fifteen years. I knew her well and trusted her completely. They were frustrated. I was frustrated too and on their behalf. The client was not feeling much better.

I was trying to express the need for a higher level path to green plan, they kept giving me the key details alongside, showing thoroughness and precision. Three times in, I still was not getting something from them I knew would have hit the mark, yet they were really trying to persuade me that their work was what was needed. I had a nagging feeling of discomfort in my chest that would not leave me.

So I asked ChatGPT to explain why I was not satisfied, biasing the prompt in favour of my team being right. I effectively asked the machine to tell me I was wrong and to let it go, but it did not. It said: the programme did not primarily have a technology problem. It had a control and translation problem. The team was telling an accurate story from entirely the wrong vantage point. They were deep inside the product, reporting from its perspective, filtering in technical detail and filtering out the human layer entirely. The two audiences who most needed to be reassured, the CEO who had to approve the release and the controllers on the ground who had to trust and use it daily, were not being spoken to at all in a way that made sense to them. Their needs, their language, their level of detail, their very existence as the people who mattered most had been quietly edited out of every document and every conversation.

My head of product management was, among other things, an experienced rock climber. Someone entirely capable of reading a wall dynamically, of staying permeable to new information as a situation unfolds, of updating a plan in real time as the environment demands it. When we were under pressure on a new business effort, it was him I turned to. And yet in this project, under the pressure of expertise and deadline and accumulated familiarity with the product, the wall had stopped being read. The plan formed at the bottom was being executed faithfully and the ability to look up and adjust had been lost.

The filter had been running without anyone choosing it. And it was quietly deciding the fate of the project.

Why the brain edits before you see

Your reticular activating system, a network of neurons at the base of the brainstem, scans the vast incoming stream of sensory and environmental information and decides, before conscious awareness has been engaged, what gets through and what does not. It is the reason you notice your name spoken across a noisy room before you have registered anything else being said. It is also the reason you can sit in the same meeting as a colleague and leave with an almost entirely different account of what happened, because your filter and theirs were set differently before you walked in.

Priming is one of the mechanisms that sets the filter. Whatever you have been thinking about, training for, or taught to look for, you will tend to find. This is the beautiful field of predictive processing, and thank goodness this mechanism is in place, because without it we would be entirely overwhelmed by the volume of incoming data our rational mind needs to process. The problem arises when it becomes invisible and unchallenged, when expertise, urgency, or collective momentum locks it onto a single frame and stops admitting new perspectives entirely.

The more experienced the team, the more deeply grooved the filter tends to be as explored in Know Too Much. See too Little from earlier on in the series. My Go-Ahead team was not making an amateur mistake. They were making an expert one, applying the most sophisticated lens they had to a problem that required a completely different lens altogether. They had stopped reading the wall.

When AI sees outside your filter

ChatGPT came to my rescue at a moment when I was getting confused by a mix of feelings, from fear of being too demanding, to an inability to articulate clearly what I needed, to loyalty to my team. I was trying to convince myself that my team was right so I could sleep better at night. But the algorithm, with its completely different set of biases, surfaced a very clear and directional solution that made me sleep that night and made the life of my team considerably easier too.

My prompt was something like: I trust my team, they are talented and experienced, they have delivered before, and yet this time the work is not landing as it should. The CEO is losing confidence. I feel we are not delivering the bird’s eye path to green we need, but perhaps I am incorrect in my assessment as I cannot fully explain why.

What came back was not just a validation of my hunch but a precise framework for what was missing. The path to green needed to be told not from the product's perspective but from the perspectives of the two audiences who mattered most: the CEO and the controllers. Each needed a different level of detail, a different language, and a different emotional reassurance. I copied and pasted the rationale directly into our email thread, making it fully transparent that a tool had come to our rescue. Within hours the team had reframed the work entirely. The client's confidence began to return almost immediately.

The reason this worked was not that AI was smarter than my team. It had a different bias in the form of no filter. It had not spent weeks inside the product. It had not absorbed the assumptions and framings that had quietly become invisible to everyone working closely on the release. It introduced a perspective from outside the filter, which is precisely what none of us inside it could generate on our own. The signal had been mine all along yet I lacked the language to make it shareable, and a tool with no stake in the existing framing to help me find it.

In this instance, AI was the different perspective that sat outside the filter and therefore saw what those inside could not. I used it to test my hunch, not to generate one. Try it, it may work for you too.

How to reset the filter before it resets you

Meditate, meditate, meditate. Research consistently shows that meditators make better decisions not because they think harder but because they are less captured by the filter already running. Specifically, they are less influenced by cognitive bias, more able to evaluate what is actually in front of them rather than what their existing frame told them to expect, and more likely to focus on the substantive merits of a situation rather than reacting to how it compares to their prior assumptions. In negotiation contexts, experienced meditators show a measurably reduced tendency to reject objectively good outcomes simply because those outcomes did not arrive in the expected form. They notice more. They filter less reflexively. And they stay permeable to what is actually there rather than what the plan told them should be there, which is in essence exactly what the rock climber needs to do on the wall.

I am not a meditation expert and found my practice and rhythm very late in life. I wish I had started when Rick Rubin did. I suspect I will write more about this and share my full journey, but for now just try to find the type of meditation that works best for you, or download Petit Bambou. It is free.

The dandelion I nearly missed

I took the photograph you see at the top of this article on a walk I almost did not take, in a moment of attention I almost did not have. A dandelion clock, perfectly backlit, a single seed still attached and on the verge of being carried away by the next breath of wind. The kind of detail that is there every day and noticed almost never.

I am reasonably certain I would have walked straight past it five years ago. Not because it was not there but because my filter was not set to concede it space. I was moving too fast, carrying too much, my attention locked onto whatever the current pressing problem was, the way it almost always was.

Meditation did not make me a different person. It changed my rhythm, my perspective, and therefore what gets through. It loosened the grip of the existing frame just enough that the quiet, peripheral, easily-missed things started arriving at conscious awareness more often. Like a dandelion seed about to fly towards its next big adventure.

What is coming next

Next week we go further into what some people see in a room that others miss entirely. Atmospheric sensing, environmental intuition, and the difference between trained attention and hypervigilance. Why walking into a space and knowing immediately what is happening beneath the surface is a trainable capability and not a personality quirk.

Next week: Reading the Room. What Some People See That Others Miss.

If this resonates, share it with someone who has ever felt that something important was missing from a conversation without being able to say what. They now have a name for the mechanism and a way to start working with it.

If you are not yet following Intuition Edge, you can do that here:

Share

Published every Monday from this week. A way to get your week started with a reminder to take a moment to train yourself to listen to better signals.

I lived in Japan in my early twenties and I have never quite stopped thinking about the quality of attention I encountered there, a willingness to sense the room before speaking in it which I quintessentially associate with Japan. I was very lucky later in my life to win some transformational work from Yamaha Music Corporation, headquartered in Hamamatsu, just a couple of hours south of Tokyo on a bullet train, and to return to this beautiful country many times over.

Subscribe now

There is a reason I keep coming back to Yamaha. Not just because of the pitch, the creative director, the client who cried, and the stomach flip that told me we had won before anyone said a word, all of which we will get to shortly, but because Yamaha as a company is built on a philosophy that the rest of this article is about. Founded in 1887 around the understanding that music making requires body, mind, and instrument to become one continuous intelligence, Yamaha has never believed in the separation as you simply cannot play well while remaining separate from what you are playing. Neither can you move an audience while remaining separate from what you are feeling. The vagus nerve, stimulated directly by music, vibration, and sound, is the highway that makes this possible, carrying signals between body, instrument, and the room music fills with emotions.

I mention this because in earlier pieces in this series I may have written about the brain and the body in a way that a brilliant MD psychologist, the husband of a good friend, gently pointed out had not quite landed the point of unity as clearly as the science actually supports. Re-reading my work, he was right. The brain and the body are one continuous intelligence that we have been describing in two parts for so long that the language itself has become deceiving. This piece is where I rectify that, with Yamaha as the starting point, because they never made the mistake in the first place.

The gut that thinks.

Your enteric nervous system, the neural network lining your digestive tract, contains approximately 500 million neurons, more than your spinal cord, operating largely independently of the brain and processing information about your environment in real time. That stomach tightening you have learned to breathe through and ignore before a slide that does not quite tell the whole story is not anxiety but data. It arrives through a different channel than the one you were trained to trust, from a system with five hundred million neurons that has been running its own analysis of every room you have ever walked into. It deserves better than being filed under nerves and moved past. Surely you do not want to impolitely ignore five hundred million of your own helpful neurons, right?

Aristotle counted five. Ward counted fifty three. Kate told me.

My friend Kate, who I first met when I moderated a panel she was speaking on, introduced me to the work of Ashley Ward, a professor of animal behaviour at the University of Sydney, whose book Sensational argues that we may have as many as fifty three senses rather than the five Aristotle identified over two thousand years ago. Aristotle’s original five were almost entirely concerned with exteroception, sensing inputs arriving from the outside world. What Ward argues, and what the science increasingly supports, is that we have many more, including those that account for interoception, the ability to sense the body’s internal state, which affect our emotions, cognition, and behaviours in ways the popular five cannot begin to account for. I do not personally agree with every classification on Ward’s list, but I am entirely with him on the broader point: there is a great deal more to train our awareness and listening skills on than most of us were ever told.

Your body sends the message. Your brain writes the story.

The vagus nerve is the carrier of that information upward, from your gut, your lungs, your heart, your skin, all the way to your brain, giving you the chance to interpret what you are receiving in the most useful way. Around eighty percent of its fibres carry information upward from the body to the brain rather than the other way around, which means the body is doing considerably more talking than most of us were ever told. And here is the part that tends to surprise people: we think we get scared and our heart starts racing. The James-Lange theory of emotion, supported by decades of subsequent interoception research, suggests it is actually the other way around. Your heart starts racing first, sends the message up, and your brain gets to decide whether to call it anxiety or excitement, two states that are, physiologically speaking, remarkably similar. What you label it changes everything. What your body did come first.

A word of caution though: this is not a licence to rebrand every difficult emotion as something more convenient. Calling low mood excitement would place you firmly in the delusional optimist camp, which is a different publication entirely.

All of which is abundantly proven by the Iowa Gambling Task we explored last week. No surprise that the best unrequested advice I felt compelled to give a good mate joining the World Series of Poker in Las Vegas in July was to work on his interoceptive abilities. But I am digressing again and back to the Yamaha story now.

The story of Make Waves.

Neil returned from his honeymoon tanned and entirely relaxed, with no idea that I had cleared his diary completely and had only three days to produce something extraordinary for one of the most important pitches of my career. The client was flying into the BBH offices in London from Hamamatsu, a company whose philosophy of embodied intelligence had been expressed through music for more than a century, and we had nothing creative to show them…yet.

Neil came up with three ideas. I do not remember the other two. When he said Make Waves, and explained how it connected music, selfhood, and society simultaneously, my stomach did the flip I described in the previous piece in this series, that unmistakable rollercoaster sensation of elation and expansion combined that I have learned over the years to trust completely. One of the clients cried when they saw the brand film Neil and the team had conjured out of thin air. Seven people in a room, a range of cultures, and one piece of knowledge that landed in all of us before a single piece of research had been commissioned.

And indeed research was commissioned and revealed that in Japan, waves carry the shadow of tsunami, and the idea needed careful brand rules to ensure it landed well in the home market. The instinct was right and the rigorous work that followed made it more universally viable, giving it the structure and the guardrail to make it travel. This is not a story about ignoring analysis but one about the right sequence: sensing first, examining second, acting with confidence third. Neil is now global creative director at Dr. Martens. Some things age very well.

What if Perplexity was there then?

I find myself genuinely curious about what would have happened had tools like ChatGPT, Perplexity, or Brandwatch been popularly available when Make Waves needed to go through global market testing. Would we have been able to write prompts clean enough of our own enthusiasm to surface an honest perspective? Would the AI have caught the tsunami association before the research did? Or would it simply have reflected back the positive bias we had already built into every question we asked? I do not know the answer. But I think the question is worth sitting with before the next time you use an AI output to validate something your body already told you it loved.

When an AI output lands and your gut stays completely flat, that flatness is information. When something tightens in your chest as you read a recommendation that your rational mind finds entirely reasonable, that tightening deserves a moment of attention before you forward it as your own conclusion.

The fluid performer.

The best decision makers, whether traders, creative directors, surgeons, or leaders, are not the ones who suppress the body’s intelligence in favour of rational analysis, nor the ones who follow every physical sensation without scrutiny. They are the ones who move fluidly between both, using analysis to examine what the body has already sensed, and using the body’s response to test what the analysis is telling them.

Part of the problem is historical. Logical analysis, the kind that can be explained in a meeting, written in a report, and measured on a dashboard, has enjoyed an extraordinary reputation since the Enlightenment and the Industrial Revolution gave us the tools to quantify almost everything. Intuition, which I would argue is not the opposite of rational thinking but another form of it, an unstructured response to pattern, experience, environmental cues, and embodied data, has been given a remarkably bad reputation at work and in life, dismissed as vague, unreliable, or simply unprofessional. It is none of those things. It is simply something harder to put in a slide or an Excel spreadsheet.

Enough of that. Time to treat both as equally worthy ingredients in service of better judgement, and to start where the evidence suggests you should: by getting better at listening to what comes from within.

What this means for you this week.

I am going to jump on the admittedly crowded bandwagon of journaling enthusiasm, possibly quite late in the game, and suggest you start an interoceptive one.

Here is how it works in practice. When you are facing a decision, simulate in your mind moving in different possible directions, and note down how your body responds to each of the options. Be as descriptive and complete as you can about the sensations you notice and where you feel them. No analysis required at this stage, no interpretation, just the raw observation, the discipline of distinguishing what you feel and where, described as accurately as you can manage.

Log it and return to it later, once the decision has played out, and map your initial bodily response against how you assess the outcome in hindsight. We do have a tendency to post-rationalise in our favour, which will make perfect sense to anyone who has read the Einstellung piece two weeks ago, but the exercise is useful nonetheless. Over time you will begin to see where your body tends to speak first and most accurately, and that is information no framework can give you in advance.

The body has been keeping better records than you have. Time to start comparing notes.

What is coming next.

Next week we explore why you notice what you notice, and why someone else in the same room notices something entirely different. The filtering mechanism behind weak signal detection, what saliency has to do with the quality of your decisions, and how AI is quietly reshaping what gets filtered in and out before you have consciously chosen to pay attention.

If this resonates, share it with someone who has ever been told their body’s response to a situation was not a relevant input. They now have five hundred million reasons to disagree.

If you are not yet following Intuition Edge, you can do that here: substack.com/@intuitionedge

Published every Friday.

You know you are treading on shaky ground when the hiring manager is looking at you with a fixed smile and a slightly lost gaze that goes somewhere behind your head, because they cannot quite look you in the face without letting off strong feelings.

Subscribe now

I did not think there was anything particularly wrong with telling someone honestly that my heart had skipped a beat five minutes into interviewing a candidate and that the feeling had stayed all the way until they left the room because they were the perfect fit. Apparently that was not good enough, and I had to go back to where exactly on the CV this or that competency was demonstrated. My heart did skip a beat when I met Elisabeth Pinto, Miles Gilder, and Cristiana Carauta, to name a few, and the fight I had to have for Cristiana was particularly memorable, as apparently her previous employer was not considered the most obvious training ground for the prestigious AKQA. She proved every sceptic wrong and went on to have a stellar career at Google and Apple after her exceptional boss left and it was not quite as much fun to stay. I am kidding. Mostly.

I should acknowledge, with the benefit of hindsight and a working knowledge of employment law in the United Kingdom, that describing my heart beating differently when I met the right candidate was not considered an entirely acceptable feedback framework by the HR department. But what if my heart already knew what a relief it would be to have competent hands arriving while the work was piling up at speed? I knew it worked. And I was right about all three of them.

What I did not have, in those interview rooms or in any of the rooms where something shifted before I could name it, was a language for what was happening or a scientific framework that would have made the HR conversation considerably less awkward. It turns out the heart is not the romantic metaphor we were sold but a precision instrument, and I can now fully use it and explain it to any HR professional who cares to listen.

Your Heart Knows First

Many cite Antonio Damasio when talking about somatic markers, and rightly so, but it is the work of the less well known Antoine Bechara that genuinely fascinated me. The Iowa Gambling Task asked participants to draw cards from four decks, two rigged to deliver occasional large rewards but consistent long term losses, and two offering smaller rewards but steady gains. Before participants could consciously identify which decks were dangerous, their skin was already registering stress responses every time their hand moved toward the bad ones. Their bodies knew the game was rigged approximately eighty cards before their conscious minds could articulate why. Some participants never consciously worked it out at all, yet their hands had long since stopped reaching for the losing decks. The body is not the passenger in decision making. It had been filing accurate reports the entire time, long before the rational mind caught up.

John Coates spent roughly twenty years on the trading floors of Goldman Sachs, Merrill Lynch, and Deutsche Bank, and what he observed there compelled him to leave and become a neuroscientist at Cambridge. The most consistently successful traders, he found, were not the ones with the most prestigious degrees or the most sophisticated predictive tools. They were the ones with the ability to sense their own heartbeat changing in real time as they approached the edge of a risk, and the willingness to act on it. The ones with the fancier credentials and the more elaborate tools turned out to be better at one thing above all: constructing persuasive explanations for why they were underperforming. Does that sound familiar?

Since Coates first published these findings in 2016, heart rate variability, the subtle variation between heartbeats, has become one of the most studied markers in performance science, associated with better cognitive flexibility, stronger emotional regulation, and more accurate judgement under pressure. It is measurable, trainable, and for most people operating in high stakes environments, almost entirely ignored. Following your heart, it turns out, can pay back in ways that show up in your pocket.

This matters more now than it ever has. When an AI tool delivers a recommendation with confidence and fluency, something in you either responds or it does not, and that response is data. If your heart rate stays completely flat when you read an output that your rational mind finds perfectly reasonable, that flatness deserves a moment of attention before you forward it as your own conclusion. AI can extend what you can see but cannot tell you what you actually sense. That distinction is the whole game.

The Heart Has Its Own Nervous System

The heart has approximately forty thousand neurons, and it sends more signals to the brain than the brain sends to it, the majority travelling upward through the vagus nerve. If you thought it was merely a pump, however important, think again. When your heart rate shifts in a meeting before you have consciously registered why, that is your cardiac nervous system processing environmental information and taking stock. The question is whether you have learned to use that information rather than dismiss it.

Think about the last time your heart quickened in a professional context before you understood why, a conversation in an interview, a negotiation, a creative session, a moment where the mood shifted without an obvious reason. What did you do with it?

What This Means For You Starting This Week

Before your next significant meeting, spend sixty seconds noticing your resting heartbeat as a baseline, not counting it, just registering it. There is a wonderful anecdote from an exhibition at the London Science Museum where an older couple were asked to participate in a live exhibit exploring how different people sense their own heartbeat without taking their pulse. The wife of forty years thought it was simply impossible. Her husband thought it was the easiest thing in the world. They were both genuinely surprised by the other’s answer, because it was something they had never once discussed in four decades together, which says something rather interesting about couples and communication as well as about the signal itself.

Try it now. Close your eyes and give yourself a minute. Can you tap your own heartbeat without taking your pulse? If the answer is no, do ten push ups and try again. I bet you can now even if the signal has not changed, simply become louder. What you need to practise is the art of tuning in, so you can hear it even when it whispers.

So during that meeting, notice when it changes and write down what was happening in the room at that moment. Do this three times and you will begin to see a pattern that has been there your entire career, quietly working for you while your mind was busy elsewhere.

That shift is data. It arrived before your analysis did. And if you find yourself trying to explain any of this to your HR department, you have my full sympathy, my complete solidarity, and the work of Antoine Bechara to cite in your defence. You may still get the fixed smile and the lost gaze, but at least now you will know exactly what your heart was trying to tell you when it happened.

What Is Coming Next

Next week we move from the heart to the rest of the body, because cardioception is only the beginning. The gut has its own brain. The skin registers decisions before the mind makes them. And there is a story about a pitch, a creative director fresh off his honeymoon, a client from Japan who cried in front of a great idea, and the moment a signal was validated by two nervous systems in two different cultures before a single piece of research had been commissioned.

Next week: Your Body Is Talking. Listen Up.

If this resonates, share it with someone who has ever been told their instincts were not a valid input. They now have the science to disagree.

If you are not yet following Intuition Edge, you can do that here: substack.com/@intuitionedge

Published every week.

And to everyone in the UK, happy bank holiday weekend. I will be camping in Scotland with no mobile phone, which feels like the most appropriate way to practise what I preach. I hope you find your own version of stillness this weekend to listen to that heartbeat.

Leave a comment

There is a phrase I have borrowed for years, from Mark Twain: it is not what you do not know that gets you into trouble. It is what you know for certain that just is not so.

I have said it in boardrooms, in strategy sessions, in conversations with leaders who were entirely convinced they were reading the situation correctly. It is the most useful single sentence I know for the moment when expertise stops being an asset and starts being a liability.

Subscribe now

Before we go further, I need to make a confession. I have always found more comfort in Socrates than in certainty, because the only thing I know is that “I know nothing” is not a sign of weakness but the most rigorous starting position available to a good thinking person.

It keeps you open not just to what you have not yet learned, but to re-examining what you think you already know. It is a form of actively practicing radical open-mindedness, so that genuinely new thinking can arrive through different context, unexpected perspective, and the kind of mistake that turns out to be serendipity. One way I have practically kept my own expertise under check, and one you can easily try in your professional or personal life, is to use syllogisms: the discipline of reaching a conclusion, looking at which premises actually support it, and testing those premises before accepting where they lead. That habit has saved me from my own pattern library many times.

And here is the tension at the heart of this article. Expertise is one of the most powerful signal amplifiers available to human judgement, the very thing that makes it possible to be rigorous about where you land and to defend that position to others with confidence. And yet, under specific conditions, it is also one of the most reliable ways to miss genuine new brilliance entirely.

Both things are entirely true. The question is knowing which one is happening to you at any given moment.

The grandmaster and the gift.

In the 1940s, a Dutch chess master named Adriaan de Groot ran a series of experiments that would take decades to fully understand. He showed chess positions to players of varying skill levels and asked them to reconstruct the positions from memory after a brief exposure. Grandmasters could do so with near-perfect accuracy, while novices could barely manage a handful of pieces. The difference was not raw intelligence but in the pattern recognition abilities built through tens of thousands of hours of absorbed experience. Where a novice saw thirty-two individual pieces, a grandmaster saw five or six meaningful configurations, instantly recognisable, immediately legible at a glance.

This is the gift of expertise. A vast internal archive of compressed experience that allows an expert to read a situation in seconds and deliver a signal that would take a novice hours of analysis to approximate. Klein’s fire lieutenant, who we met in article two, was running exactly this system. So was Gottman, reading a marriage in fifteen minutes with ninety-five percent accuracy. So were the Getty art historians who felt the wrongness of the kouros before they could name it, as we explored in article three.

Expertise, at its best, is the mechanism by which experience becomes sensing. That is genuinely extraordinary, and worth protecting.

But there is a shadow side. And it has a name.

The Einstellung effect. Or: how the answer you already know stops you finding a better one.

In the same era, the psychologist Abraham Luchins demonstrated something that should unsettle every expert reading this. He gave participants a series of water jar problems requiring them to measure specific volumes using jars of different sizes. The early problems could all be solved using the same reliable method. Participants learned it and moved on. Then Luchins introduced a problem that could be solved using the familiar method, but could also be solved using a much simpler approach. Most participants never saw the simpler solution. Not because they lacked the intelligence to find it, but because their brain, having found a working pattern, had stopped looking for alternatives.

He called this the Einstellung effect, which means “set effect” in German. Trust the Germans to find the most efficient word for the moment your brain stops being efficient. It is one of the most well-documented phenomena in cognitive science, and it operates most powerfully in people with the most experience, precisely because they have the most deeply grooved pattern libraries to default to.

Here is where it gets interesting. Timothy Wilson and Jonathan Schooler at the University of Virginia ran a study in 1991 asking people to rate strawberry jams. Non-experts gave ratings that closely matched those of professional food critics. Then a second group was asked to explain their reasoning before rating. Their scores fell apart entirely, diverging sharply from the expert consensus. The act of verbalising what they sensed corrupted the signal. The experts, by contrast, had the language to articulate their judgement without losing it.

Does any of this feel familiar? Think of the last time you were asked to justify a decision over and over, at work, in a relationship, or by an overbearing parent, in a world where “because I just know” is not acceptable currency in a professional environment or apparently at Sunday lunch either.

Rationalisation has its place, and a genuinely useful one. The problem is not the tool. It is reaching for it before the signal has had a chance to land.

The practical tip is simple: when you sense something clearly, note it down before you start explaining it to anyone. The act of writing it in its raw form preserves the signal before verbalisation has a chance to corrupt it.

The lesson cuts both ways. The novice's clean, unmediated signal is genuinely valuable and genuinely fragile. The expert's language protects the signal, but the pattern library can also override it. Knowing which is happening in any given moment is the name of the game.

What the beginner sees that the expert misses.

Shunryu Suzuki, the Japanese Zen teacher who brought beginner’s mind into Western consciousness, put it plainly: in the beginner’s mind there are many possibilities, but in the expert’s mind there are few.

My living proof of this is from 2007, when someone was courageous enough to believe in me leading a major global automotive account through the most important re-launch in the brand’s history — the iconic Fiat 500 — without knowing that I did not yet know what the word fleet meant as I had never led any client in mobility. I ended up leading teams that won what was at the time the highest concentration of innovation awards I had ever been part of, from CNET recognition to a Grand Prix at Cannes Lions to best digital piece of the decade from Contagious magazine. I was asking genuinely stupid questions. Most of them were, in fact, stupid. But that tenth question, the one that came from not knowing what the situation was supposed to look like, opened up a category-first way of thinking that nobody inside the industry had found, because everyone inside already knew how things were supposed to look. My naivety was my asset. My expertise was what allowed me to fence off the doubters with a sensible and compelling story. A brilliant team, a great deal of grit, and a certain amount of luck did the rest.

The practical implication is not to abandon expertise. It is to cultivate what researchers now call actively open-minded thinking: treating your own existing knowledge as a hypothesis rather than a conclusion, and asking whether it is still the best available explanation for what you are observing, or whether you are pattern-matching your way to a familiar answer because it is faster and more comfortable than sitting with genuine uncertainty.

Choose deliberately between expertise and naivety. Only you know which one the moment is asking for.

Socrates was practicing this two and a half thousand years before cognitive science gave it a name.

Now add AI. And watch the pattern library get complicated.

Assisted intelligence tools introduce a version of the Einstellung effect that operates at a different scale. When an AI output arrives fluently and confidently, the expert’s pattern recognition system compares it against its own library, finds a reasonable match, and relaxes. The question worth building as a habit is not does this output look right, but does it look right because it is right, or because it fits a pattern I already hold?

The practice moment.

Before your next significant decision, ask yourself one question: am I seeing what is actually here, or am I seeing the version of this that fits what I think I already know?

Write down one genuine anomaly. That note is worth more than it looks. It is the beginning of seeing clearly.

And if this piece made you think of someone whose pattern library you trust completely and occasionally want to throw out of a window, send it to them. They will know exactly what you mean.

If you are not yet following Intuition Edge, you can do that here: substack.com/@intuitionedge

Published every week.

Last week I wrote about Klein’s fire lieutenant who ordered his crew out of a burning building seconds before the floor collapsed. He could not explain why, and in the aftermath he wondered whether something supernatural had guided him. It had not.

This week the aim is to go back to that moment, not to retell it but to open it up, because what happened to him is happening to you every day. In smaller ways and lower stakes, in every meeting, every conversation, every decision you make where time is short and information is incomplete.

Subscribe now

Once you understand the mechanism behind it, you will realise it has been shaping your judgement your entire life, quietly, continuously, and without ever asking your permission. That same mechanism, is the one that reads a failing marriage, detects a ten million dollar fake, and tells you something is wrong with a deal before the numbers have given you a reason to say so. It is called subconscious pattern recognition, and it is only the first of many mechanisms that make intuition work. The most accepted, the most researched, and it explains a great deal yet it is only the beginning. Hence the series.

Your Brain Is Running a Programme You Cannot Access.

In 1983, art dealer Gianfranco Becchina approached the J. Paul Getty Museum in California with what he claimed was a rare ancient Greek marble statue, a kouros from the sixth century BC. There are fewer than two hundred such statues in existence, most of them partially broken or retrieved in fragments. This one was almost perfectly preserved.

The team at Getty did everything right. They took the statue on loan and spent fourteen months authenticating it, analysing the marble using electron microscopes, mass spectrometry, and X-ray diffraction, tracing its provenance through legal documentation, and running every scientific test available to them. The science confirmed it. The documents all checked out. The Getty bought it for just under ten million dollars.

Then the experts arrived.

Evelyn Harrison, one of the world’s foremost authorities on Greek sculpture, felt something the moment the cloth was pulled back, though she could not immediately say what. Federico Zeri, an Italian art historian on the Getty’s board of trustees, found himself staring at the statue’s fingernails, unsettled by something he could not name. Thomas Hoving, former director of the Metropolitan Museum of Art in New York, always made a note of the first word that came into his mind when he encountered something new. When he saw the kouros, that word was fresh. Which seems to me not the word that should occur to you when looking at a two thousand year old statue. Georgios Dontas, head of the Archaeological Society of Athens, described feeling an invisible wall between himself and the sculpture.

None of them could articulate why. None of them needed to. They all knew.

The statue was a fake, and it took years of further investigation to prove what those experts had registered in seconds. You now get why this story made such a powerful opening for Malcolm Gladwell’s Blink. In two seconds of looking, Harrison, Zeri, and Hoving understood more about the essence of the statue than the Getty team had understood in fourteen months. And where, you might reasonably ask, were all these experts when they were actually needed, before the purchase occurred? Jokes aside, what Gladwell identifies here is something genuinely powerful. He calls it the adaptive unconscious. The part of the brain that processes information rapidly and automatically, drawing on everything we have learned and experienced, and delivering conclusions as feelings, hunches, or an immediate sense of rightness or wrongness, long before our conscious mind has had time to form a coherent thought about what it is experiencing.

Your Brain Is Running a Programme You Cannot Access. And That Is a Very Good Thing.

Leading neuroscientists are clear on this: if you had to consciously process every piece of information your brain receives right now, you would not be able to move. The volume of sensory data arriving every second, from your environment, your body, your memory, and your social context, is so vast that conscious awareness can only handle a tiny fraction of it. Your subconscious does the rest.

Subconscious pattern recognition is what happens next. It is the mechanism by which all of that absorbed experience gets organised, compared, and delivered back to you as a signal. When something matches a known pattern, you feel a sense of rightness. When something deviates from it, you feel friction, a sensation that something is off even when you cannot yet name what.

The art historians at the Getty were not guessing. They were pattern matching. Decades of studying Greek sculpture had built an internal library so detailed and so refined that a single glance at a fake triggered an immediate mismatch signal. Their nervous systems detected it before their conscious minds could form a sentence about it.

Gary Klein spent decades studying this mechanism in firefighters, military commanders, intensive care nurses, and chess grandmasters, and what he found was consistent across every domain. Expert decision makers do not analyse their way to conclusions but they rather recognise it. They walk into a situation and their brain, running its pattern matching programme at extraordinary speed, delivers a signal, sometimes a feeling of rightness, sometimes wrongness, sometimes a pull toward a particular course of action that feels almost inevitable. Klein called this Recognition Primed Decision (RPD) making, and the insight at its core is deceptively simple. The expert does not generate options and compare them. They sense the right move first, then run a quick mental simulation to check whether it will work, and if it does, they act. The pattern comes before the analysis. The signal arrives before the sentence.

This is not a shortcut. It is a more sophisticated form of processing than conscious analysis alone allows, faster, more integrative, and in domains where genuine expertise exists, often more accurate.

From Love to Death. Why the Stakes Range Does Not Matter but the Mechanism Does.

The fire lieutenant’s story is dramatic because the stakes were absolute. But subconscious pattern recognition is not reserved for burning buildings.

John Gottman is a psychologist at the University of Washington who has spent decades studying what makes marriages succeed or fail, bringing more than three thousand couples into his laboratory and watching them talk, not about the big things but about the small ones. A disagreement about a dog. A conversation about weekend plans. From one hour of watching a couple interact, Gottman can predict with ninety five percent accuracy whether they will still be married fifteen years later, and from fifteen minutes his accuracy is around ninety percent. One of his colleagues can do it in three.

He is not guessing. He is thin-slicing, reading the patterns in how two people talk to each other, the micro expressions, the quality of attention, the presence or absence of contempt, and matching them against an internal library built from thousands of hours of observation. The single biggest predictor he found is not conflict or disagreement but contempt, a pattern so specific and so embedded in the way certain couples interact that it shows up whether they are arguing about something serious or debating what to have for dinner.

From predicting whether a marriage will survive to knowing when a building is about to collapse, the mechanism is identical. Pattern recognition built through experience, operating below conscious awareness, delivering its conclusions as a signal before the conscious mind has caught up. It is the thing that makes you pause before clicking a link that looks almost right but not quite, the feeling that a deal is too good before you have done the maths, the sense that someone is performing warmth rather than feeling it before you have any evidence to support the suspicion, the moment you walk into a house and know immediately it is not yours even though it ticks every box on the list.

It is also behind the signal you have been ignoring. The one that has been telling you for months that the job no longer fits, that you keep overriding with a very reasonable list of financial justifications and a deep reluctance to disappoint your dad. Your nervous system has been reading the mismatch between who you are and where you are, and filing its report faithfully every single day. The question is never whether the signal is there. It always is. The question is what you do with it.

Fast or Slow. The Signal Decides, Not You.

And here is where it gets nuanced, because the answer is not always to act. The fire lieutenant had seconds, and the signal demanded immediate movement. His pattern recognition system delivered its conclusion and the only intelligent response was to trust it and move. Dyson’s signal, by contrast, the one that kept him at the workbench through 5.127 prototypes, demanded the opposite. It required him to hold the signal over years, to let it inform rather than dictate, to integrate it with each failed experiment and each new data point until the solution finally emerged. And the Getty art historians felt their signal the moment the cloth came off the statue, but what followed was not immediate action. It was the beginning of a long process of examination, questioning, and investigation that eventually proved what the signal had already told them.

Speed of response is not the point. The signal tells you something is true, but it does not always tell you what to do next or how fast to do it. Sometimes it requires you to act before language arrives, and sometimes it requires you to pause, sit with the discomfort, bring in other information, sleep on it, and let the full picture emerge before you move. And sometimes, on reflection, the signal turns out to be conditioning or fear dressed up as knowing, and the wisest thing you can do is nothing at all. The skill is not in trusting the signal blindly but in learning to read what it is asking of you, and that is something that develops with practice, with the right prompts, and with the willingness to examine rather than simply obey.

What This Means When AI Is in the Room.

Here is a practical question worth sitting with. When an AI tool delivers a recommendation, a forecast, or a structured analysis, do you accept it because it is right, or because it is fluent? Because it arrived with confidence and structure and something in you relaxed at the sight of it?

That relaxation is worth examining. Your pattern recognition system responds to AI outputs the same way it responds to any other input. It is comparing what it receives against what it knows. If the output matches your experience, you will likely feel at ease. If something is off, even subtly, you may feel friction before you can name why, and that friction deserves attention before you commit to the output as your own conclusion.

The people who tend to use AI most effectively are not the ones who trust it most readily. They are the ones who notice their own signal in response to what it produces. They use AI to extend what they can see, then bring their own pattern recognition to bear on what it shows them. The signal does not compete with the output but rather it interrogates it. That combination, AI augmenting what you can access and trained human sensing applied to what it surfaces, is what better decisions are actually made of. Not one or the other. Both, working together, with you in the chair.

What Is Coming Next.

Next week we move from the mind to the body. Because the signal does not only live in your pattern library or on the screen where the answer readily appears. It lives in your chest, your gut, your breath, and the subtle sensations you have spent years learning to ignore. And it turns out those sensations are data too.

That mechanism is called interoception. And it is entirely learnable.

If this resonates, share it with someone who has ever dismissed a signal as just a feeling. That signal had a source that deserves better than dismissal.

Follow Intuition Edge: substack.com/@intuitionedge

Livia Bernardini. Leadership advisor, author, and builder of The Intuitive Advantage.