The Competitive Advantage Is Shifting

In my previous article, I talked about how AI is reducing the cost of writing software.

Because of this, other skills are becoming more important: networking, audience building, marketing, distribution.

But there is another skill that matters even more now: domain expertise.

Because if everybody can build software, then the competitive advantage changes.

We are moving from:

“Can you build it?”

to:

“Can you build the right thing?”

And those are completely different skills.

The bottleneck is shifting away from pure technical execution and toward: knowledge, experience, expertise,intuition, instinct.

Especially inside niche industries.

The AI has knowledge about these domains. Sometimes surprisingly good knowledge.

But from my experience, it still lacks the depth required to build truly effective software for real users inside specialized industries.

Because expertise is not just information.

It is context. It is experience. It is pattern recognition. It is instinct. It is understanding what actually matters and what does not.

And LLMs cannot provide this by themselves.

They still require the right person asking the right questions, providing the right context, identifying the right problem, and guiding the system in the right direction.

This is where humans still create enormous value.

Another skill that becomes critical in this new era is the ability to search.

Not simply searching Google.

I mean: the ability to find the right information at the right time for the right audience to solve the right problem.

This skill was always important. But AI amplifies its importance dramatically.

Today, information is everywhere. You can ask almost anything to an LLM and receive an answer instantly.

So the value is no longer: “Do you know information?”

The value becomes: 

• Can you navigate information effectively?

• Can you identify the real problem?

• Can you filter noise?

• Can you recognize bad advice?

• Can you connect the right information to the right context?

And honestly, this is why I think most “prompt engineering” advice online is overrated.

The wording matters less than people think.

What matters more is:

• the context,

• the expertise,

• the intuition,

• the understanding of the problem itself.

Experienced people ask better questions because they understand the domain better.

And this leads me to a question.

If searching, navigating information, and solving the right problems are becoming critical skills…

Why are companies still recruiting like it’s 1995?

Why are interviews still heavily based on memorization?

Why are developers still rejected because they forgot some advanced technical detail or a best practice during an interview?

In real life, engineering is rarely about memorizing information.

It is about:

• identifying problems,

• finding the right information,

• evaluating solutions,

• and solving the problem effectively for the customer and the business.

So why are we still mostly assessing memory instead of problem-solving ability in the AI era?

I honestly think this gap will become increasingly visible over the next few years.