Working With the Layer of Intelligence That Makes Change Last

Your nervous system might have a different opinion to you.

Over time, working with people, I’ve noticed something fairly consistent. The shifts that really last don’t usually come from insight alone. They tend to happen earlier than language, earlier than reasoning. Long before someone can explain why they feel anxious, stuck, or on edge, their nervous system has already made a call about what’s safe and what isn’t. More and more, that’s where I find myself working. Not by arguing with it, but by offering information it can actually use and responding to what it shows me in return.

An example of the nervous system working and predicting at this deeper layer can be seen very early on, even in babies. One well known example from psychology is object permanence. At a certain age, babies realise that something still exists even when it’s out of sight. We know this before they can tell us, not because they explain it, but because their nervous system reacts when expectations are broken. In research, this shows up through measures such as galvanic skin response. When something doesn’t behave the way the system predicts, there’s a clear physiological signal of surprise.

This kind of learning happens without words. The nervous system updates its expectations based on what actually occurs, not on what’s explained to it.

I see the same process playing out later in life. Emotional responses are learned in much the same way. The nervous system builds expectations about danger, safety, and what needs to be prepared for. Those expectations can remain long after the original situation has passed, because from the system’s point of view, the response once made sense.

Over time, this has changed how I think about intelligence itself. I no longer see it as a single thing that lives only in our thoughts. I see multiple systems within us, each with their own role and goals, often operating without language. More recently, I came across the work of Michael Levin, a leading developmental biologist who studies how intelligent behaviour can be influenced in systems that don’t rely on a brain or language.

In some of his work, this involves looking at how to influence the behaviour of cells that are doing something unhelpful, including cancer cells. Rather than trying to destroy them outright, the question becomes whether their behaviour can be redirected by changing the signals they respond to. The focus is on persuasion rather than force, and on working with the goals the system already has rather than fighting against them. That way of thinking maps closely onto how I work with emotional responses.

When I’m working with someone, I’m not trying to override these systems or get rid of responses. I assume they’re there for a reason. Instead, I offer information and alternative viewpoints that might help the system reassess what it’s protecting against. The aim isn’t persuasion in the usual sense. It’s to provide new data that could help the system feel safe enough to adjust a pre-learned response.

For an update to take effect, the system has to be in a state where change is possible. There are brief windows where the nervous system becomes more flexible and able to revise what it’s learned. Part of my work is helping to create those openings, then working within them rather than trying to force anything outside of them.

Once we’re working within one of those windows, I think about what follows in terms of telemetry. By telemetry, I mean live feedback from the nervous system about whether it’s open to this specific information we’re offering it for consideration. That feedback guides what happens next. If the system isn’t open, there’s a reason. That reason tells us what information is missing or what concern still needs addressing. If the system is open, we know an update is possible, and when that happens the change can occur in seconds.

To make this more concrete, the information I’m talking about isn’t generic reassurance. It’s specific to the response that’s running.

For example, when someone experiences flashbacks linked to a painful event from the past, the nervous system is often behaving as if that danger is still present. If we can offer clear, present day information about safety and if the nervous system feels safe enough to accept it and actually indicate this through feedback, the flashback response can switch off very quickly. When that update happens, it can feel almost like flicking a switch.

Another example might be crippling guilt about a past action. In these cases, the nervous system may be using guilt as a way of preventing repetition through punishment. I might offer the distinction between guilt and regret, where regret keeps the lesson learned but removes the ongoing self punishment. If the system agrees that this still serves its protective goal, the guilt response can drop away just as suddenly.

In both cases, the words are simply proposals. The deciding factor is whether the nervous system agrees that the new information allows it to do its job more effectively.

Over time, working with so many people, I’ve also learned something else. While people are endlessly individual, the nervous system tends to operate within the same logic from person to person. It tends to rely on a small number of predictable strategies, each with its own blind spots. Certain types of information reliably land, and certain assumptions reliably keep responses running. Once you recognise one of those patterns and what unlocks it, the same principle often applies elsewhere. That doesn’t mean applying a formula. It means knowing what this kind of system usually needs to hear in order to stand down.

When this happens, it doesn’t usually feel dramatic, but the speed of the shift can catch people by surprise. There’s no forcing and no trying to calm anything down. Often, people notice they feel quite different. Almost as if a weight has been lifted and it happens immediately. Certain problems seem to fade or disappear on their own, and their thinking becomes calmer and clearer without trying.

I don’t measure galvanic skin response in the room. I don’t need machines to work this way. Over time, I’ve developed quick, internal, and very streamlined ways of interacting at this pre-linguistic level and checking how the system is responding.

Working this way has changed how I think about change itself. I’m less interested in effort and more interested in whether the right information has reached the right level. When it has, the system does the rest. My role is simply to notice, respond, and stay out of the way once the update has happened.