Hello!
It’s Thursday, 26th March 2026. Hello and welcome back to Bold Efforts! Last week I wrote about safe jobs. I have been thinking about that phrase since then, and I think it points in the wrong direction.
The question is not which jobs are safe. The question is what becomes valuable when intelligence itself becomes cheap.
For a long time, a big part of white-collar work was simply the cost of producing a decent answer. A decent deck. A decent memo. A decent market map. A decent draft. A decent design direction. A decent first pass on a problem.
That work took time because human attention was the bottleneck.
You felt this most inside large organizations. Five years ago, I could spend three months pushing through a small design change in a product at an enterprise. There were workshops, revisions, dependencies, alignments, handoffs, comments, more comments, and then another round because someone senior wanted to revisit the logic. Today I can overhaul the design in a day, with a tightly defined design system and the right AI tools.
This is a real shift. It is not just about speed. It is about the falling price of competence. And when the price of competence falls, the whole hierarchy around it starts to move.
Many people still talk about AI as if the main question is whether it can do your job. That is too blunt. A machine does not need to do all of your job to change the economics of your job. It only needs to make an important part of it cheaper, faster, and easier to access.
That is what is happening now. Intelligence, at least in its operational form, is becoming abundant.
You can get decent copy in seconds. Decent research in minutes. Decent designs in an afternoon. Decent analysis without hiring a team. The floor is rising very quickly.
But when the floor rises, the question changes. If good-enough output is easy, then what exactly are people paying for?
I think the answer is trust. Not trust as a slogan. Trust as economic value.
Who do I trust to define the real problem? Who do I trust to filter signal from noise? Who do I trust to make the tradeoff when every option looks plausible? Who do I trust when the stakes are high and the decision will have consequences? Who do I trust to say no?
This is the part that gets missed. AI is making intelligence cheaper. But it is not making judgment cheap. In some ways, it is doing the opposite.
When anyone can generate ten options, the person who matters is the one who can tell which option is wrong. When anyone can create a polished deck, the person who matters is the one who can tell whether the strategy underneath it makes sense. When anyone can produce a clean design, the person who matters is the one who understands the user, the constraints, the tradeoffs, and the business well enough to choose the right one.
This is why so much AI output feels impressive at first and hollow a few minutes later. It often answers the visible question. But real work is usually hidden one layer deeper.
The visible question is what should we build. The real question is which problem is important enough to deserve a build.
The visible question is which candidate looks strong. The real question is who you would trust in a messy room six months from now.
The visible question is what does the market want. The real question is which demand is durable and which signal is just noise wearing a suit.
This is why I think the premium layer of work is shifting. For years, many people were rewarded for producing intelligence. Now more people will be rewarded for directing it, editing it, challenging it, and standing behind it. That changes careers in a serious way.
If your value is mostly that you can produce competent output, you are in a vulnerable position. AI is coming directly for that layer.
If your value is that you can reduce uncertainty for other people, you are moving closer to the scarce layer. That is what a trusted person does.
A trusted operator brings calm to complexity. A trusted manager makes the right tradeoff earlier. A trusted advisor sees second-order effects before they become visible. A trusted founder knows which feature, hire, or market to ignore. A trusted doctor, lawyer, investor, or executive carries responsibility, not just information.
This is why I do not think the future belongs only to the best prompt writers or the most enthusiastic users of AI tools. That will matter, of course. But it is not enough.
The real winners will be people whose judgment becomes easier to see because AI is everywhere.
When everyone has access to powerful tools, your edge is no longer access. Your edge is discernment. Your edge is taste. Your edge is being right when it counts. Your edge is being the person others are comfortable betting on.
This also explains a deeper tension in the labor market. As intelligence gets cheaper, companies will need fewer people for some kinds of work. That part is real.
But the work that remains will not simply be more complex. It will be more exposed. More visible. More tied to consequences.
In a world flooded with generated output, people will pay more for people who can absorb risk. That is what trust really is. It is not likability. It is not polish. It is not confidence theater. It is credible judgment under real conditions.
And credible judgment is expensive because it takes time to build and reality to test. You cannot fake it for long.
A person can sound smart in one meeting. A person can look sharp in one memo. A person can hide behind good formatting for a while. But over time, reality runs the audit. The people who can be trusted keep separating themselves from the people who only look fluent.
I think that is the real story behind this moment. AI is not just automating tasks. It is exposing where the real value was all along.
A lot of modern work was built around the production of intelligence because intelligence was expensive. Now that it is getting cheaper, the market starts paying more attention to the next layer up.
Who can judge. Who can decide. Who can take responsibility. Who can be trusted. That is why I would rather ask this than ask which jobs are safe.
When intelligence becomes cheap, why should anyone trust you with something important? That is the harder question.
It is also the more useful one. See you next week. Thank you for reading.
Best,
Kartik
I write Bold Efforts every week to think clearly about where work and life are actually headed, not where headlines say they are. If you want these essays in your inbox, you can subscribe here.