The Score and the Music

The Score and the Music

I want to start with a debate …. and a disagreement.

Not a healthcare debate, although I think it has quite a lot to say about healthcare.

This debate took place in Paris, in April 1922, between two of the most famous intellectual figures in the world.

One was Albert Einstein, whose theories of relativity had transformed modern physics.

The other was Henri Bergson, at the time probably the most famous philosopher in the world, known particularly for his thinking about time, perception, memory and human experience.

The room was packed yet hushed, and what followed became one of – if not the - most famous encounters between science and philosophy in the twentieth century.

It is sometimes simplified into Einstein saying one thing about time and Bergson another, with Einstein winning because, well, it was his event - Bergson was originally an audience member.

No It was more interesting than that.

Bergson was not saying Einstein was wrong. He was saying that the time measured by a clock is not the whole of the time we live.

Einstein gave us time that could be measured precisely - in seconds, intervals, synchronised clocks, and the time needed to calculate how bodies - move through space.

But Bergson was interested in a different sense of time. Lived time, psychological time and how that differs in the situations we experience.

I think of the distinction like this.

Einstein gives us the notes on the musical score - their position, their interval, their timing.

Whereas Bergson gives us the melody we hear - we experience - when those notes are played.

Yes the score is a real thing. And so is the music.

But history nevertheless remembered the debate as a victory for Einstein, and the century that followed gave enormous authority to a mode of knowing built on observation, measurement, instrumentation and representation.

So why this preamble? I want to suggest today that something of that distinction between Einstein and Bergson remains with us in healthcare.

I describe it as the difference between an observed world and a lived world.

The observed world is what we can detect, measure, classify and represent.

The lived world is where life is actually happening.

And healthcare has become extraordinarily good at the first.

From that observed world have come remarkable advances in our understanding of physiology and disease, instruments that sense, detect, diagnose and monitor change in the body at extraordinary precision, and diagnostic frameworks for hundreds of conditions through which health systems assess risk and intervene.

From that observed world have come remarkable advances in our understanding of physiology and disease, instruments that sense, detect, diagnose and monitor change in the body at extraordinary precision, and diagnostic frameworks for hundreds of conditions through which health systems assess risk and intervene.

I recently tried to count the physiological measures now available through different forms of remote patient monitoring.

I stopped after forty.

And around those sit all the other things we measure …

Clinical outcomes, utilisation, complications, prevention, quality of life, patient-reported outcomes, device performance, adherence, activation, self-management.

This world is real, and it matters enormously. Your evidence lives here. Your regulators live here. Your reimbursement arguments live here.

I am not dismissing any of it. But I do want us to notice something.

As the instruments through which we observe health become more sophisticated, the things those instruments can see acquire more and more authority over what counts as important.

And the better we become at seeing within that frame, the easier it becomes to mistake the frame for the world.

So we have an interesting situation.

We have never possessed more means of seeing the patient.

And yet an enormous part of the patient's life remains largely outside our field of vision.

Let me make that concrete with the actual experience of Ruth. She is a composite, but there is nothing that unusual in any part of her.

Her numbers are good, therapy is working, readings stable, and she is adherent.

Nothing in her data would cause her clinical team any concern.

On every term through which the care system currently encounters her, she is doing well.

But Ruth no longer travels anywhere unless she knows exactly where she can stop. Her partner has started checking on her several times a day. She has stopped accepting invitations, because the uncertainty is exhausting.

Her mornings take twice as long as they used to, because of a sequence of checks and preparations that nobody has ever asked her about.

She sleeps badly. She plans constantly.

In short, she is clinically stable yet Experientially contracting

And nothing has gone wrong with the monitoring for that to be true. The data is accurate. What is missing is not accuracy but part of reality.

Because beneath everything we measure is the world in which the condition is actually lived.

The body, of course, but also the other conditions within that body. Relationships. Work. Money. Memory. Confidence. The home. The workplace. The places somebody goes and the places they have quietly stopped going.

The routines they develop. What they expect from tomorrow and what they no longer expect at all.

The condition does not remain politely inside the clinical category we have given it. But moves into the rest of life.

Over the past decade we have been developing a framework for understanding this world which we call Real Lived Experience, or RLX.

It has grown out of work across very different contexts: chronic pain among military veterans, post-conflict experience in Northern Ireland, obesity, anxiety, ostomy care, ageing, autism and chronic wounds, together with a deeper engagement with philosophies of perception, embodiment and relation.

RLX works across five interacting domains of living experience.

But the important thing isn't really the number five.

It is what happens between them.

A change in the body changes movement. Altered movement changes where somebody goes. Where somebody can go changes their social relations. Those changing relations affect confidence, mood, attention, memory and expectation. Those changes alter how a treatment or device is used. And that feeds back into the body.

So experience is not simply a collection of variables.

It moves.

And it moves differently for different people.

What RLX gives us, then, is something closer to a living field: a continuously changing composition of body, perception, social relations, environments, technologies and resources.

And once you begin to look from there, other things become visible.

You notice not simply symptoms, but what somebody is trying to preserve, recover, avoid or make possible. We call those Real Lived Outcomes.

You see the workarounds people invent which never reach a medical record because nobody thought to ask

You see the burdens of vigilance and attention,

and the hidden costs which later become costs to the system.

And critically, you begin to see movement.

Whether experience is holding.

Whether something new is becoming possible.

Whether life is becoming precarious.

OR Whether it is getting stuck.

And once we begin to see this wider world, another question appears.

What, ultimately, are we trying to create through long-term-condition care?

Which brings me to the idea of what we call, livability.

By livability I mean whether, and how much, a life can proceed, and potentially expand, with a long-term condition.

That means proceeding safely, but also increasingly free from the unnecessary attentional work of the condition

the checking, remembering, anticipating, organising and adapting through which a condition can progressively occupy more and more of somebody's world.

And critically, it means doing so while care remains available when care is needed.

That last part matters.

Livability is not independence, and it is not telling people to cope better so the system can withdraw.

It is creating the conditions in which a life no longer has to remain organised around the condition.

And shifts the question about what long term condition care must do to …

Can the life proceed?

And what does it currently cost that person to keep it proceeding?

Consider Ruth.

She would score reasonably well on most quality-of-life instruments.

What those instruments will not show you is how much of her day the condition now claims, or that the radius of her life has been contracting for two years.

And so Livability is less a score of a quality of life than a way of understanding its possibility and movement.

And when you start from there, some familiar healthcare ideas begin to look slightly different.

Clinical management extends into the wider conditions that allow a life to work. We remain interested in the disease, and become equally interested in what allows somebody to live alongside it.

For adherence, instead of asking how we make somebody comply more reliably, we ask how we could reduce the vigilance and effort the treatment requires of them in the first place.

For Product performance, that becomes more than whether a device works under defined conditions. It becomes how well it fits inside an actual life – taking account of a person with other conditions, a working day, an ageing body, a particular home, a particular way of moving.

And preventing complications - while remaining essential - becomes attached to a larger ambition, that of enabling somebody, eventually, to live past the condition rather than simply becoming highly accomplished at managing it.

So while our usual concerns haven't disappeared – it’s just that the field around them has widened.

Think about two devices.

They may perform identically against every clinical endpoint yet one asks a person to check it four times a day and the other tells them only when something has changed.

And so in effect, the same performance … but a different amount of a life.

All this raises the question of how we reach this lived experienced world.

We have traditionally done that through interviews, ethnography and, increasingly, patient participation. All of these are valuable, and we use them …. but they have limits.

Ethnography enters a life for a period and then leaves.

Interviews and surveys depend on what somebody can recall and recognise as relevant.

Participation brings people into our processes, while their contribution is still typically translated back into the concepts we use.

And importantly, some of the most consequential aspects of living experience are precisely the things people don't or can’t even articulate, because they have become normal, because they seem too trivial to mention, or because nobody has ever asked a question that would make them relevant.

So the challenge is not simply to ask people more questions.

It is to develop a greater capacity to perceive.

That is the challenge our technology work is trying to address.

We have built Bergson.ai as a platform around the RLX framework.

At enterprise level it works across any condition and population, for research, innovation, design, evidence and strategy. Give it a condition, a population and a setting, and it returns what is likely to be present in the lived world around the clinical picture

Including the burdens people carry, the conditions producing them, the relations a life is held together by, the real lived outcomes they are working towards, and where they have become stuck.

Importantly. everything it returns is a candidate, and not an answer. It does not tell us or predict what anyone experiences, and it is no substitute for human engagement. Building something that claimed to .. would reproduce exactly the problem I have been describing.

But it does widen the field before the engagement happens, so that we arrive with knowledge, perspective and questions that were not available to us before, and then we check those with the people and contexts they describe.

And we are now extending the same architecture to the point of care, across different setting types, so that the field widens inside the encounter itself.

And then there is Ooex, originally developed with armed-forces veterans living with chronic pain as the experiences are for here.

It begins from the other side of the problem: the fact that almost all chronic illness is lived outside the clinical encounter. It captures meaningful events and changes in experience over time, between and beyond appointments, during the thousands of hours in which life is actually happening.

What emerges from Ooex is not another sequence of scores, but trajectories. Whether of a life sustaining itself, something becoming possible again, or a manageable situation beginning to contract.

Which brings me to one final idea, because I think there is another reason why this matters, and it extends beyond healthcare.

I call it attenuation.

By attenuation I mean the progressive contraction of what can become consequential within a clinical or device or monitoring encounter –

Where fewer aspects of the real situation retain the capacity to shape what happens next, even while the encounter itself remains active, connected or increasingly intensive.

Here is an example from a study published just last month – a review of an emerging evidence base.

Forty per cent of GPs say they have used an ambient AI scribe.

A University of Edinburgh review of twenty-seven studies found that these tools are being judged almost entirely on documentation time and workload, while they also change the consultation itself.

Aspects such as facial expression, gesture, emotional state and the patient's own account of their illness are the parts most likely to be left out.

Yes … the encounter is more densely connected than ever … more of it is recorded, but less of it can travel.

And once the AI summary becomes the record, what did not survive the consultation is less likely to shape the next one.

That is attenuation. It’s not the disappearance of any clinical encounter but a contraction in what the encounter can respond to.

And I think this presents an important warning for integrated monitoring and therapy.

Because we could quite easily create an extraordinarily intensive ongoing technological encounter with a person.

One involving continuous sensing, transmission, analysis and feedback.

And yet though, at the same time, progressively narrow the range of that person's reality to which the system is capable of responding.

In effect, we become more connected while becoming less responsive.

The system becomes increasingly present in the person's life while the person's life becomes increasingly absent from the system.

I will repeat that - the system becomes increasingly present in the person's life while the person's life becomes increasingly absent from the system.

And perhaps that is one of the risks of an attenuated healthcare future. Indeed, of an increasingly attenuated society.

We are becoming extraordinarily good at making life legible through data, classifications, algorithms, predictions and models.

But the particular person, the unexpected, the contradictory, the emergent, the things that do not fit our concepts easily, can become less and less consequential …. precisely as our systems become more sophisticated.

Which takes me back to where I started.

There is a view gaining ground, in some of the most serious boardrooms in technology, that life is fundamentally a form of computation.

I understand its appeal …. but I think it is wrong.

A description of the functions and processes within a life is not the same thing as the life.

The danger is that once we say life is computation, it becomes imaginable that a sufficiently powerful description might eventually exhaust the living thing.

That is the moment the score claims to be the music.

My own vision of AI in healthcare and elsewhere is almost the reverse.

The challenge ahead is not simply one of ever greater artificial intelligence.

It is one of experiential intelligence.

By that I mean the capacity of people, organisations and technologies to become responsive to more of the living reality they encounter.

What is particular to this person, in this place, at this moment.

What has not yet become a variable. What may be changing before anybody has found the words for it.

And that doesn't require us to choose Bergson instead of Einstein.

We need Einstein. We need measurement, precision, instruments, and computation that detects patterns no human being could perceive unaided.

But we need Bergson too. We need duration, movement, emergence, difference and relation, and an acknowledgement that life exceeds the instruments through which we try to know it.

And so in a sense, Einstein tunes the instruments, holds and turns the pages of the score.

But Bergson plays the melody.

Or in our terms, Clinical intelligence helps us understand what is happening to the condition.

Experiential intelligence helps us perceive what is happening to the life.

And Ruth needs both. But she has only ever been offered one.

So Einstein and Bergson. Albert and Henri.

That's a double act I'd pay to see.