Mindset, Skillset, Toolset

By mid-morning, Samantha was deep in a feedback loop.
.
On her desk sat a battered notebook with a few phrases circled in red, a laptop running BrandGPT, a shared prompt library, and an analytics dashboard that kept flickering with new readouts. She started the morning by testing three simple brand criteria:
.
“fast, flexible, human”
.
BrandGPT responded by producing forty headline variants in a few seconds. The analytics harness ranked them according to an evaluation framework based on past customer interaction data, and five clear outliers rose to the top. Samantha read those outliers slowly, felt which ones landed the wrong way on tone, and then rewrote the prompt. The next batch came back noticeably closer to what she wanted.

Was Samantha doing the work? What was really “thinking” in that moment was Samantha’s whole system: her brain, her body, the notebook, the prompts, the analytics, and the model. The intelligence emerged in the way all of these components were arranged.

The Extended Mind, in Real Life

We often talk as if “mind” is something that lives entirely inside the skull. In practice, our thinking has always spilled out into our bodies and our surroundings. Cognitive scientists sometimes call this the “extended mind:” the idea that cognition is distributed across internal memory and intuition, bodily states and sensations, and all the external scaffolds and tools we use to support our thinking.

You can see this in the simplest everyday example. You write a to-do list: instead of holding everything in your head, you let the paper remember for you. When you glance at the list later, your brain instantly recovers the priorities you decided on earlier. The list is a small piece of your memory and executive function that you’ve pushed outside your head.

The same thing happens when you talk through a hard problem with a colleague, sketch on a whiteboard, or rearrange sticky notes on a wall. Your environment becomes part of how you think. Your body is part of it too: the way you gesture when you explain something, the sense of “gut feeling” when a decision is wrong, the restless tap of your foot when you’re stuck and trying to unconsciously shake loose a new idea.

Generative AI tools are the brand new addition to this extended system. They don’t sit apart from your mind; they can become part of your cognitive process, offering pattern recall, rapid variation, simulation, and critique. Used well, they don’t replace your judgment but give it more reach. Used poorly, they become noise that pulls you away from your own values and sense of quality.

How Generative AI Fits Into This System

Once you see work as an extended cognitive system, generative AI becomes less threatening and more legible. It is not an all-knowing oracle. It is more like a very fast, reasonably smart colleague with no values, no skin in the game, and no real understanding but also with incredible recall and an ability to remix patterns at scale.

One way to use AI in this system is to treat it as an idea generator while you remain the editor. Imagine you’re drafting a strategy memo. You begin by clarifying the audience, the constraints, and the core message. Then you ask the model to suggest ten possible structures and five very different openings. The model offers a range which may include some that are predictable, and a few surprising. You scan them quickly and feel which ones resonate with your goals and your voice. You might combine the framing of one with the sequencing of another. The model expands the space of possibilities; your judgment selects and shapes.

Another role is that of simulator, where the AI acts as a rehearsal room. You paste in a draft policy and ask the model to respond as a frontline employee, a skeptical middle manager, and a privacy lawyer. The responses are not perfect, but they surface plausible questions, concerns, and edge cases. You refine the policy in response. You remain the policy maker, deciding what tradeoffs to accept. The model simply accelerates your ability to see around corners.

AI can also serve as a translator. Suppose you have a single core idea—say, that your company is shifting from measuring production to measuring computation as a primary source of value. You articulate the idea once, in your own words. Then you ask the model to help you adapt it into a short LinkedIn post, a three-slide executive summary, and a two-minute spoken script. The underlying thinking is yours; the model gives it different shapes suitable for different channels.

Finally, AI can act as a mirror. You might feed it a dozen emails you’ve written and ask for patterns in your communication style. It may notice, for example, that your messages are very clear but often too long, or that your tone skews more urgent than you intend. It might offer concrete suggestions for making your writing more concise or more empathetic. The model, in this role, reflects you back to yourself and helps you see habits that are otherwise invisible from the inside.

In all of these cases, the AI is not the mind; it is a component in a larger system that includes your knowledge, your values, your body, and your environment.

Your Body Still Matters

When people talk about “brain + AI,” the body tends to disappear from the picture, but it remains a crucial part of cognition. The way you sit, breathe, move, and gesture changes how you think. Experiments show that people solve problems differently when they are allowed to move their hands, draw, or physically rearrange objects. The body is not just carrying the brain around; it’s actively helping structure thought.

Notice what happens to your own body when you feel like you’re competing with AI. For many people, there’s a subtle tightening: shoulders creep up, breathing becomes shallow, scrolling speeds up, and attention feels scattered. The work shifts into a defensive mode: “I need to prove I can do this better than the machine.” In that state, judgment suffers. You skim instead of read, you react instead of reflect, you rush past the quiet sense that something isn’t right.

Now contrast that with the state you’re in when you treat the system—your brain, your body, your external scaffolds, and the AI tools—as a single extended mind. You sit up, but not tensely. You give yourself a moment to think before firing another prompt. You pay attention to the small signals of “this feels off” or “this is promising.” You allow the model to generate possibilities, but you do not outsource your values or your standards.