You probably live most of your life on autopilot. This isn’t a judgement or a criticism, but an observation about how most people seem to function.

The autopilot is a common, ordinary state of consciousness where you don’t notice that you’re auto-approving your familiar, habitual responses to stimuli. More than a state of consciousness, it’s a state of bodymind, since body and mind are not separate. The autopilot disables your capacity to consent to your own actions.

Picture the scene. You’re walking to the kitchen to get a glass of water and you see a cookie on the counter. It occurs to you that you would enjoy eating the cookie, and immediately you reach for it and put it in your mouth. You may even forget about the glass of water.

At what point in that chain did you decide to eat the cookie? Did you decide at all, or was your behaviour automatic?

I’m writing this in October 2026, when referencing Claude Code’s --dangerously-skip-permissions flag isn’t yet charmingly outdated. That phrase means that Claude will do whatever it decides to do without stopping to ask you first. When using an AI, you would consciously decide to switch that mode on (or perhaps you’ve used it enough that you do that on autopilot as well).

Your autopilot doesn’t have an equivalent ‘on’ button, because over the course of your life it has become a default setting. It gets triggered whenever you decide to do something: sit, speak, type, anything.

Returning to the cookie, you probably didn’t consciously decide to eat the cookie; at least, the moment when choice was possible was so brief and subtle that you didn’t really have a chance to think and give or withhold consent. Something in you decided for you and you didn’t realise you could say no.

The autopilot is always chasing a prediction of what your goal will feel like when you get there. More fundamentally, the autopilot wants the felt experience of familiarity.

When you eat the cookie, you may expect it to feel enjoyable or satisfying. But if eating a cookie is your habit, then you’ll also expect it to feel familiar, while choosing not to eat it will feel unfamiliar.

The problem, though, is that familiarity isn’t something you really feel as familiarity. Instead, familiarity feels right, while unfamiliarity feels wrong. What this means is that you default to doing what feels right, rather than doing what you set out to do, which was to get just a glass of water.

In this example, I’m just talking about eating a cookie, but this dynamic happens all the time, like when you reach for your phone, snap at a loved one or agree to another useless meeting. Of course, sometimes you decide to look at your phone, but many other times you just find yourself scrolling.

The phenomenology of the autopilot

In general, it’s easiest to notice the autopilot after it has disengaged, hard but possible to notice it before it engages, and nearly impossible to notice it while it’s running.

You may wake up from a trance of scrolling and then realise that you weren’t there for it. You may sometimes notice the urge to reach for your phone, before or as your hand starts to move. But once you’re on autopilot, it usually takes some new stimulus, internal or external, to break the spell. Either that, or you reach the autopilot’s goal and it releases you by itself.

To help make sense of why this is, let me introduce my Four Spectra of Presence model. These spectra are expressions of four aspects of your experience:

• Knowing. Knowing is the quality of knowing that you know or being aware that you’re aware. When knowing is high, you’re there for your experience. When knowing is low, you’ve checked out.
• Awareness. Awareness is the field in which you can notice things in you and in the world around you. When your awareness is expanded, the whole world is there. When it’s contracted, everything but what’s got your attention goes away.
• Attention. Attention is the ‘spotlight’ that moves around within awareness based on what’s important to you, or to a part of you. When attention is grippy, it’s difficult to move it away from whatever it’s on. When attention is free, it can move easily, even if it doesn’t.
• Availability. Availability is a felt dynamism in response to the world’s invitations to act. When you’re available, there is no transition between the idea of action and action itself. When you’re unavailable, there’s a sense of rigidity and the need to remove a kind of brake before action is possible. In the past, I have referred to this as couldness.

Presence is a state with high knowing, expanded awareness, free attention and unlocked availability from which the world seems brighter and closer, you function better, and you can make new, conscious choices.

The autopilot is the opposite. Its defining feature is that knowing goes offline, so there’s no one there to notice what’s happening or to make a new choice. At the same time, the autopilot tends to push the other spectra away from presence as well: your awareness contracts, your attention gets stuck and your bodymind becomes rigid.

When you’re on autopilot your body also works in an autopilot-shaped way, with excess muscle tension, gripping and an overall pattern of pulling yourself down physically. When the autopilot triggers, you respond habitually, which means that everything you do will feel right.

The autopilot and habitual action go hand in hand, so feeling right is another indicator of the autopilot. However, feeling right isn’t salient, so it simply doesn’t register at all. Nothing about your experience on autopilot feels wrong or noteworthy.

This explains why it’s nearly impossible to notice the autopilot while it’s running, and why it’s easiest to notice after it releases you. There’s a distinct experience of those spectra moving back towards presence: ‘you’ come back, the world comes back, your attention releases, and some previously unseen brake disengages.

It’s difficult to notice the autopilot as it starts to come online because of the speed at which it happens. Even a small dimming of knowing and you’re gone, though it gets easier to notice and prevent the autopilot with practice.

Freedom to change

Noticing the autopilot, and learning to prevent it from taking over, gives you agency. Or, rather, it gives you the capacity not to do whatever your habits decide for you and to land in a place where new options are available.

As long as you are unable to withhold consent from your familiar, habitual patterns, they are all you will ever be able to follow. This matters because the world you’re responding to is complex and constantly changing (and so are you).

Unconsciously repeating a learned library of fixed habits would be rational under the assumption that the world is simple, static and predictable, but the world has never been that. In fact, living from that assumption means that you’re responding to your image of the world rather than to the world itself.

And since your habits are so familiar, and the autopilot dims your ability to notice that it even exists, you rarely get any signal that there is a gap at all. You can only respond to what you notice.

The way to get out of this trap is to come to a complete stop. The moment you stop trying to get somewhere is the moment you wake up to the world as it is, unsimplified.

It may feel wrong, but that doesn’t make it so. As a general rule, your feelings tell you what’s familiar, not what’s true. From unfamiliarity you can finally act in a new way, because anything truly new must feel that way.

The cost of learning to prevent the autopilot is greater responsibility. The more you are there for your life, the more you are responsible for the choices you make. On autopilot, you could honestly say you didn’t know what you were doing. Once you’ve noticed that you have a choice, you can’t.

Your bodymind will still sometimes --dangerously-skip-permissions and go on autopilot, as high stakes and deep habit will fire it despite your best intentions. The real practice isn’t to avoid ever going on autopilot, but to notice without judgement when it happens and carry on.