AI didn't make you replaceable

Before we dive in, two resources worth bookmarking:

Conversations about AI and work tend to follow a predictable pattern. I've seen this where someone publishes a list of the jobs most at risk of automation. People scroll through it quickly, privately relieved when they don't see their role on the list. Then the next article comes out, the list is slightly different, and the whole cycle repeats.

That conversation is too blunt. The better question isn’t whether AI can replicate your job description. It’s whether it can replicate you.

A job description is a list of tasks. Many of them, at enough scale and with enough training data, can be approximated by a machine that never needs a lunch break or a moment of self-doubt. The sharper question is whether AI can replicate the way your mind works. That means the questions you form, the things you notice, and the judgements you make in the room that nobody asked for but that turned out to matter enormously.

For a lot of introverts, the answer is that AI can try and imitate this badly, and not for the reasons people often assume.

ChatGPT is a superpower if you know how to use it correctly.

Discover how HubSpot's guide to AI can elevate both your productivity and creativity to get more things done.

Learn to automate tasks, enhance decision-making, and foster innovation with the power of AI.

Download the free guide

AI is producing a kind of professional homogenisation in companies. When everyone uses the same tools to draft their messages, summarise their thinking and prep for meetings, the outputs start to look the same.

This is typically an output that's clean and competent, but more importantly, difficult to attribute to a certain individual. Research from Cambridge has already identified this risk in hiring. Calibrated on existing employees, the tools surface candidates who look most like whoever is already there. That narrows the range of thinking that gets into the building rather than broadening it. The same logic applies inside organisations once people are hired. AI tools are useful, as they save time, reduce friction, and improve first drafts. On the flip side, they also tend to smooth out the qualities that make someone's thinking different from another's.

Distinctiveness, it turns out, is one of the few professional assets that is genuinely hard to replicate at scale. Pew Research Center found that 52% of workers are worried about the future impact of AI in the workplace. The data misses a subtler anxiety.

AI may not take your job, but it can make your distinctiveness harder to see. For introverts who have spent years building credibility through the quality of their thinking rather than the volume of their presence, this concern is already practical. It's already happening in organisations that have adopted AI tools without thinking carefully about what gets lost when the rough edges are automated away.

Consider what AI is bad at, and what many introvert strengths sit directly on top of. AI can't sit with ambiguity without being asked to resolve it. It can't notice that the answer to the question in the room isn't the question that needs asking. It can't sense that the team's mood has shifted, making the planned decision premature. It can't hold an unpopular position unless it's told to. Those things are decision skills.

They are high-value professional contributions that look quiet from the outside and generate disproportionate results when the people doing them are given the right conditions. Adam Grant, Francesca Gino and David Hofmann found that introverted leaders can drive stronger group performance when their teams are already proactive. Harvard Business School unpacked the finding in its piece on why introverts make the best leaders for proactive employees.

The reason matters. Introverted leaders can be more receptive to initiative from others, and they're less likely to dominate the room before the best thinking has had a chance to surface. That quality is very important in my view. As AI starts to handle an increasing share of the obvious work, the less obvious work becomes more valuable. I mean human judgement, and the willingness to say the thing nobody wants to hear in a meeting that has already made up its mind.

For introverts, the risk lies in visibility. AI may replace the visible parts of other people's output while the invisible parts of your own contribution remain just as invisible as they always were. It stays unattributed, unnoticed, and gradually discounted in a world that is getting better at rewarding what is loud and legible. Which means the question isn't whether your thinking is good enough. It almost certainly is. The question is whether enough of the right people can see it.

Avoiding AI tools is the wrong lesson here, and I'm not going to suggest that. They're useful, and the people ignoring them entirely are looking to solve the wrong problem. The point is to stay visible inside your own output, and to make sure the things that are specifically yours don't get smoothed away in the process.

Before using AI to draft anything that carries your name or your judgement, write one paragraph in your own words first. Even if it's a rough draft or you write it badly. That paragraph is your thinking, and everything you do to that afterward is editing. Skip that step, hand the blank page straight to the machine, and you've outsourced the most valuable part: your own starting position.

AI is good at clean answers. It's worse at showing the reasoning that led someone with your experience to one. The thing you noticed that the data didn't surface, the risk you mentioned that everyone else was avoiding, and the question you asked that changed the meeting's direction. Write that down in the follow-up notes, in your messages app, and in the brief email after the decision's made. For example: "We chose Option B because the risk on Option A isn't visible in the numbers yet." The purpose here is attribution, showing that your thinking was in the room, not just your attendance.

The qualities that make introverts valuable in an AI-homogenised workplace are also the qualities that an AI efficiency culture can erode first. The willingness to sit with a problem before resolving it, the resistance to premature consensus, or the habit of noticing what's missing rather than only what is present. Slowness gets reframed as a problem to fix, and discomfort gets automated away. Guard both.

The evidence here is still developing, so I wouldn't overstate it. Early findings on AI tools, cognitive offloading and critical thinking carry a useful warning, echoed in the MIT Media Lab preprint Your Brain on ChatGPT. Lean too heavily on AI for thinking, and the skills you stop using may weaken.

That should matter to anyone whose professional edge depends on judgement. The introvert who is slower to delegate their thinking to a machine may be preserving something that becomes competitively important once the novelty of AI-generated smoothness wears off.

If you are an extrovert in an AI-enabled workplace, you've got a problem that may be different from the one you expected. The tools that make you faster also make you less distinctive. When your first draft is AI-assisted, and your colleague's first draft is AI-assisted, and your competitor's pitch is AI-assisted, the differentiator is no longer speed or confidence in the room.

The differentiator becomes what you think, separate from what the tool suggested you think. That is where your quieter colleagues have been sitting this whole time. Think of the person on your team who has something specific and unpopular to say, who reaches for the thought before the tool, who sounds less polished but means more.

In an AI-homogenised world, that person is an asset. They don't need coaching into sounding like everyone else. Give them the room to be heard before the agenda moves on, and the meeting will be better for it, and so will the decision.

If you are trying to work out whether AI is helping you think or quietly replacing the thinking:

The aim is to find where your fingerprints are, and where they've started to fade.

Try these three things this week without abandoning your AI tools.

The Politeness Tax

Why being easy to work with can quietly cost you.

The anxiety about AI replacing workers is real, and in some industries it's already well-founded. I'm not going to pretend that every quiet introvert with good judgement will sail through the next decade unaffected. But the version of the AI story that says the future belongs entirely to the loudest and fastest confident people in the room is too narrow.

The work that matters over the next ten years is the work AI can't do cleanly. That means holding a position under pressure, noticing what's missing from a confident consensus, and asking the question that slows everything down long enough to stop an expensive mistake. It also means building the kind of trust that comes not from a polished email, but from someone who shows up consistently and means what they say.

Introverts have already been doing that work for years (mostly without credit), in rooms that weren't built for them. AI has made that easier to see, but it's also made the people who were never really thinking that deeply slightly easier to spot.

In your corner,
— Steven

P.S. Someone you work with is doing their sharpest thinking somewhere nobody can see it. Send this their way.

P.P.S. Strong work that blends into everyone else's stays invisible no matter how good it is. From Invisible to Unforgettable is the free guide for changing that.