The Artisan and the Algorithm

There was a time when I could drive across a city and remember how to get back. Not perfectly. Not with turn-by-turn precision. But with a sense of direction — landmarks, instincts, a mental map built slowly with every wrong turn and correction.

Today, I can get anywhere faster than ever before. And yet, I have no idea where I am. Somewhere along the way, I stopped navigating. I started following.

Google Maps didn’t just change how I travel. It changed what my brain practices.

And that’s the uncomfortable question I’ve been thinking about:

If we outsource navigation long enough, we lose the ability to navigate.

What happens when we start outsourcing thinking?

This isn’t a new fear. Every generation has worried that its tools were making it weaker. Socrates warned that writing would destroy memory — that humans would stop knowing things and start merely referencing them. He was partly right. And yet, what followed made everything else possible.

The printing press threatened the scribes. The calculator threatened mathematicians. The internet threatened — well, everything, depending on who you asked.

Each time, the technology won. Each time, humanity adapted. And each time, something was lost and something larger was gained.

But this moment is different.

Every prior technology extended what the human body could do — move faster, communicate further, calculate more accurately. AI is the first technology that reaches into how the human mind works. It doesn’t just improve the output. It offers to bypass the process by which the output is understood.

And it is doing so at a speed that gives us almost no time to think about what we’re agreeing to.

The telephone took 75 years to reach half the world. The internet took 20. The smartphone took 10. ChatGPT reached 100 million users in two months — faster than any technology in history.

What took generations to distribute now takes weeks. What took years to learn is now available instantly.

Which means the decisions individuals, teams, and organizations make today about how they relate to AI will compound in ways that are difficult to reverse. The mental habits being formed now — by students, analysts, doctors, lawyers, writers — will calcify into institutional norms within a decade.

This is not a moment for panic. It is a reason for clarity.

The wrong question

The conversation consuming boardrooms, career anxieties, and headlines today is simple: Will AI take my job?

It is an important question. It is also the wrong one.

The more honest observation — the one hiding underneath the jobs debate — is this: AI hasn’t primarily destroyed jobs in its first wave. It has changed what competence looks like.

A mediocre analyst with AI can now produce what once required a good analyst.

A good analyst can produce what once required a team.

An exceptional analyst can do things that were not previously possible at all.

The threat isn’t replacement. It’s invisibility.

The professional who uses AI to confirm what they already believe, who lets it draft the memo they would have wrestled with, who reaches for the summary before forming a view — that person is still employed, still respected, still productive.

Their decline is quiet.

It compounds slowly.

And by the time it is visible, it is difficult to reverse.

And this is where something genuinely interesting is happening — a divide most organizations haven’t fully reckoned with yet.

Consider two people sitting at the same table.

The first is twenty-two. They have grown up with the internet, with search, with tools that make information instantaneous. AI is not a disruption to them — it is an environment. They are fluent in it the way a native speaker is fluent in a language: without translation, without friction, without ego.

They can summon frameworks, drafts, analyses, and synthesises faster than any generation before them. Intelligence, in a very real sense, is available to them on demand.

But access to intelligence is not the same as understanding it.

The second has twenty years of experience. They have seen cycles. They have been wrong in high-stakes situations and learned things that cannot be fully written down. They know where models break. They know the difference between an answer that is technically correct and one that is actually true.

They have built something that has no analogue in any training data: judgment earned from failure — the kind of judgment that exists only because it was once wrong.

The current narrative puts these two people in competition. The young person’s fluency threatens the older professional’s relevance. The experienced person’s seniority resists the younger one’s energy.

That framing is not just wrong. It is wasteful.

Because the most valuable thing in the age of AI isn’t fluency with the tool. It isn’t experience without the tool. It is what happens when the two are combined — deliberately, structurally, with clarity about what each brings that the other cannot replicate.

The artisan and the algorithm

Here is the simplest way I know to say it.

AI is extraordinarily good at the articulable — everything that can be written down, structured, trained on. It synthesizes faster than any human. It never forgets. It does not tire.

On the terrain of the known, it is close to unbeatable.

But human expertise — especially expertise earned over years — lives largely in the inarticulate.

The judgment that comes from having been wrong in a situation that mattered. The instinct that fires before reasoning catches up. The knowledge of where the map ends and the territory begins.

AI can generate answers. It cannot tell you when the question itself is wrong.

AI makes the articulable cheap. Which means it makes the inarticulate more valuable than it has ever been.

Think about what this looks like in practice.

A young doctor with AI can synthesize years of literature on a rare condition in minutes. She can flag diagnoses that might otherwise be missed. She can cross-reference drug interactions instantly. These are real gains. Lives will be saved by them.

But the experienced clinician knows something the model doesn’t. He has seen the pattern fail. He recognizes when something feels slightly off — not because he can fully explain it, but because he has lived through it. That knowledge is not in any database. It lives in him.

Together, they produce something neither could alone.

The same dynamic plays out in law. Junior lawyers with AI can draft, surface precedents, and stress-test arguments at remarkable speed. But senior lawyers know which clauses will actually matter — not in theory, but in this context, with this counterparty, in this moment.

In investing, the gap is starkest.

AI can screen businesses, model scenarios, and synthesize years of data. But the most important information about a business rarely appears in filings. It appears in how a founder responds when challenged. In what is not said. In the subtle shift in tone over time.

That pattern recognition is not algorithmic. It is earned.

What connects these examples is structure.

In every field, there is knowledge that is transmissible — it can be written down, taught, trained on. And there is knowledge that is not — it can only be accumulated through experience.

For most of history, the first layer was expensive to access. The second was the natural byproduct of time.

AI has made the first layer nearly free. Which means organizations built on controlling access to information have lost their moat almost overnight.

But the second layer — judgment — has never been more scarce. Or more valuable.

The twenty-two year old and the twenty-year veteran are not competitors. They are, if designed correctly, the most powerful combination in the room.

One brings the tool.

The other brings the judgment to know when the tool is right, when it is wrong, and when the question itself needs to change.

Who has the right to win

Most organizations are asking the wrong question about AI.

They ask: how much should we use? How fast should we adopt? Which tools should we implement?

These are implementation questions.

The real question is structural: What kind of organization are you building — one that uses AI to accelerate its best thinking, or one that uses AI to replace thinking it was never doing well?

The answer will determine who survives the next decade.

The organizations that struggle will not be the ones that resist AI. Resistance is temporary. They will be the ones that adopt AI without understanding what they are trading away.

They hire for credentials rather than judgment. Their hiring systems were built to certify people who passed the right filters. AI can pass most of those filters now. If that is your selection mechanism, you are selecting for the wrong thing.

They have hierarchy that slows decision-making. AI compresses the time between question and answer. An organization that requires five approval layers to act on an insight will be consistently outmaneuvered by one that can move at the speed the tool enables.

They treat AI as a department instead of infrastructure — building a priesthood and leaving everyone else as passive consumers of outputs they cannot evaluate.

And their senior people stop being curious. That is not wisdom. That is calcification wearing the costume of authority.

The organizations that thrive look different.

They pair deliberately. The junior teaches tool fluency upward. The senior teaches judgment downward. Both directions simultaneously, as a designed feature of how work gets done.

They protect the formation of thought. AI enters after thinking has begun — not before. They understand that the unassisted thinking that precedes the tool is where real judgment lives.

They hire for judgment under uncertainty, not output quality. In a world where AI makes outputs cheap, the scarce resource is the ability to recognize when the output is wrong.

And their leaders model intellectual humility visibly. Because a senior person who updates their view because of something a junior colleague surfaced — publicly, without defensiveness — creates a culture that no policy document can replicate. That single behavior, repeated consistently, is worth more than any AI strategy presentation.

The right to win belongs to organizations that understand one thing clearly: AI has changed the cost of information and the speed of analysis. It has not changed what good judgment is, where it comes from, or how long it takes to build. The firms that treat these as the same problem will optimize their way into irrelevance. The ones that treat them as complementary — and design accordingly — will compound in ways their competitors won’t be able to explain.

I still use Google Maps every day. I am not going back to navigating without it. And I am not suggesting anyone else should. But I have started, occasionally, turning it off. Taking a route I half-remember. Getting it slightly wrong. Correcting. Building the map again. Not because the map doesn’t work. Because I do not want to forget that I can make one.

The question for every individual, every team, and every organization is simple:

Are you following, or are you navigating?

The tool is not the problem. It never was.

The problem is forgetting that you were the one who was supposed to know where you were going.

Disclaimer:

The performance-related information provided in this newsletter/blog is not verified by SEBI. The content is intended solely for internal circulation and general informational purposes. It does not constitute investment advice or any form of financial recommendation.

The research information shared herein may contain inaccuracies or typographical errors. All liability for actions taken or not taken based on the content of this newsletter/blog is expressly disclaimed.

No reader, user, or browser of this Newsletter / blog should act or refrain from acting based on any information in this newsletter/blog without seeking independent financial advice. Use of, and access to, this publication or any links or resources provided within do not establish a portfolio manager-client relationship between the reader, user, or browser and the authors, contributors or Itus Capital.