What does it mean to be AI-native rather than AI-enabled?

Being AI-native means AI is the default starting point for every task, not a tool you reach for when you hit a wall. An AI-enabled worker drafts the old way and pastes into a model to polish at the end. An AI-native worker opens the model first, frames the problem with it, and rebuilds the workflow around it. The difference is the starting point, and that starting point is what now separates professionals who stay employable from those who quietly age out of the market.

How does a K-shaped skill profile differ from a T-shaped one?

A T-shaped profile is deep in one discipline and broadly aware of the rest, which was the standard career advice for years. A K-shaped profile is good at many things rather than the single master of one, with several real peaks of ability. As AI merges roles like design, front end, data, and back end into fewer seats, the K-shaped professional who moves across all of them survives the merge. The T-shaped specialist who only goes deep in one lane is the one most exposed to being cut.

Why has experience become a career liability in the AI era?

Experience becomes a liability when it hardens into resistance. Senior professionals who insist on their old workflow, such as drawing every layout by hand, are now being let go for refusing to adapt rather than for poor results. The layoffs that once hit junior teams have moved up to leadership level. Seniority makes the resistance more expensive, not safer. The fix is not less experience, it is the willingness to feel like a beginner again and rebuild the work around AI.

When should a professional start paying for AI tools out of their own pocket?

Now, rather than waiting for an employer to provide a licence. Frontier tools cost about twenty dollars a month, roughly two thousand rupees, which is five coffees at a Starbucks. Subscribe on a personal device, evaluate the field monthly, and switch when something better appears. Company restrictions are an excuse, not a ceiling. The point is not the feature, it is buying yourself a new default so AI becomes the place your work begins instead of a tool you have to request.

What is the first step to becoming AI-native?

The first step is to audit where AI sits in your last five tasks. For each one, note whether you opened AI first or reached for it only at the end. If most were end-of-task polish, you are AI-enabled, not AI-native. Then pick one real task and rebuild it from the model outward, starting the work inside AI rather than bolting it on. Most people skip this audit because it stings, but that sting is the honest signal that the workflow has not actually changed yet.

AI-Enabled Is a Career Risk. AI-Native Is the Job Now.

Being AI-enabled is a career risk because it means you reach for AI only when you hit a wall. AI-native means AI is the default starting point for every task you touch. That gap, between reaching and starting, is the line that now decides who stays employable and who quietly ages out. I spent two days last weekend correcting people who thought they were already on the right side of it.

The correction that carried the whole weekend

I was running two live training sessions for early-career professionals at IIT Madras Pravartak, a UI and UX course built around agentic AI. The richest material did not come from my slides. It came from the open question and answer that opened each morning.

On the second day a participant introduced himself with some pride. He said he was "AI-enabled." I stopped him on the spot. No, I said. You need to be AI-native. AI-enabled means you reach for a tool when a need shows up. AI-native means AI is where you begin, before the need is even framed. You do not bolt AI onto your old workflow. You rebuild the workflow around it.

He looked a little stung. Most of the room did. Because almost everyone in that hall, students and working professionals alike, had quietly placed themselves in the "enabled" bucket and assumed that was the finish line. It is not the finish line. It is the trap. The enabled professional still opens the same file, still drafts the same way, still treats AI as a fancy autocomplete they summon at the end. The native professional has already changed the shape of the work.

I pushed the point a little further because the room needed it. Being enabled feels like progress, and that feeling is exactly what makes it dangerous. You learn three prompts, you save twenty minutes a day, and you tell yourself you have adapted. You have not. You have decorated the old process. The task still starts in your head the way it always did, and AI arrives late, as a helper you call in when the work is mostly done. That sequencing is the whole problem. The person who starts with the model is not getting a faster version of the same job. They are getting a different job, one where the first move is a conversation that surfaces options you would never have written down on your own.

I asked the room a simple question to test where they actually sat. When you began your last piece of work, what did you open first? For almost everyone the honest answer was a blank document, a design file, a spreadsheet. The model came later, if at all. That is the tell. Native is not about how much AI you use across a week. It is about what you reach for in the first thirty seconds of a new task.

That one correction is the entire argument of this essay, so let me make the stakes plain before I hand you the framework.

Experience has quietly become a liability

I have watched a pattern repeat at the senior end of the market that should worry anyone who leans on their years. Experience, the thing we were all told to accumulate, has become a liability in a specific way. Senior people resist. "I will still open Figma and draw my boxes by hand." That sentence used to sound like craftsmanship. Now it sounds like a resignation letter written in slow motion.

The numbers around this are not gentle. The layoffs that used to land on junior teams have moved up the chart to leadership level. In the United States the so called silver tsunami is real, with experienced leaders being let go not for poor results but for refusing to adapt to a tooling change. The resistance is the risk. The seniority makes it more expensive, not safer.

I say this as someone who is on the wrong side of any youth advantage. I did a commerce degree, no engineering. My first job was at a McDonald's in Bandra from 1997 to 2000, and I did every role from cleaning the floor upward. Somebody from a hospitality counter is now advising the largest companies in the world on how to use AI. I tell every room the same thing. Do not get disheartened by your starting point. Get disheartened by your refusal to start over.

What protects you is not the years on your CV. It is whether you are still willing to be bad at something in public. Every senior person I have watched come through this transition well has one trait in common. They sat down at the new tool and let themselves be clumsy with it for a few weeks, in front of people who used to look to them for answers. That is the real cost, and it has nothing to do with intelligence. The leaders who get cut are rarely the ones who could not learn the tool. They are the ones who could not stand to look like a learner again. Seniority makes that feeling sharper, because the higher you sit, the more it costs your ego to be the slowest person in the room for a month. Pay that cost anyway. It is the cheapest insurance you will ever buy on your own career.

The framework: the four shifts from enabled to native

If "be AI-native" is the instruction, the obvious question is how. I gave the room four shifts. Each one is a habit, not a tool. You can adopt all four this quarter without anyone's permission.

1. Start every task with AI, not at the wall. The enabled worker drafts alone, then pastes into a model to polish. The native worker opens the conversation first, thinks out loud with the model, and uses it to frame the problem before producing anything. The output improves, but that is the smaller gain. The bigger gain is that your judgment gets trained against a fast collaborator every single day.

2. Build a K-shaped skill profile, not a T. We were all taught to be T-shaped, deep in one discipline and broadly aware of the rest. T-shaped is finished. The shape you want now is K, like your grandfather's old door key with several peaks along one edge. Good at many things, not the single master of one. The corrected proverb fits here. Jack of all trades, master of none, is still better than master of one. The roles are merging in front of us anyway, so the person who can move across design, front end, data, and back end in one seat is the one who survives the merge.

3. Pay for your own tools. Stop waiting for your employer to hand you a licence. The frontier tools cost about twenty dollars a month, which is five coffees at a Starbucks, roughly two thousand rupees. Subscribe on your personal device, evaluate the field every month, and switch when something better appears. Company restrictions are not your ceiling. They are your excuse.

4. Ship public proof, not a CV. I told the designers in the room a hard line. Behance is dead. Show me your GitHub. Stop describing yourself as a designer or a developer and start pointing at deployed, original work in a public repository. Not a fork of someone else's project. Your own thing, live, with a link. Put a QR code to a working build where your resume used to sit. A claim on a CV is a promise. A working link is a receipt. In a market where anyone can list the same five tools under skills, the receipt is the only line that survives a five-minute review by someone who actually knows the work.

Enabled                         Native
   |                              |
reaches for AI    -->     starts with AI
T-shaped depth    -->     K-shaped range
waits for tools   -->     pays for tools
describes work    -->     ships public proof

The four shifts are cheap. None of them needs a budget approval or a new title. What they need is the willingness to feel like a beginner again, which is exactly the thing experience makes harder. Notice that none of the four is a tool you have to master. Tools change every quarter, and chasing the newest one is its own kind of trap. The shifts are habits about where work begins, how broad you grow, who pays for your edge, and what you can point at. Get the habits right and the specific tool stops mattering, because you will swap it out the moment a better one lands. That is the quiet advantage of treating this as behaviour rather than software. Behaviour compounds. A licence expires.

Three things to do this week

Audit your last five tasks for the starting point. For each one, write down whether you opened AI first or reached for it at the end. If four of the five were end-of-task polish, you are enabled, not native. Expect this to sting a little. That sting is the signal.
Subscribe to one frontier tool on your personal device. Spend the two thousand rupees this week, not next quarter when procurement gets around to it. Pick one real task and rebuild it from the model outward. You are buying a new default, not a feature.
Push one original thing to a public repository. Not a tutorial clone, not a fork. A small, deployed, working artefact with a link you would actually share. By Friday you should have one URL that proves you ship, not just talk.

What to read next

While you are here, the back catalogue has more on the human side of this shift:

About

Anees Merchant writes one essay every Tuesday about enterprise AI, agentic systems, and the human side of the work. He is the author of Merchants of AI, a TEDx speaker, and a doctoral researcher in Human-AI Communication at the Swiss School of Business and Management.

The newsletter version, with extra commentary, goes out separately on Mondays. Subscribe at /newsletter.

See you Tuesday.