We have misunderstood what AI is replacing

For the past two years, most conversations around artificial intelligence have focused on replacement. Will developers be replaced? Will recruiters become obsolete? Will lawyers need fewer junior associates? The assumption behind all these questions is the same: if AI can perform a task faster, cheaper and more consistenly, then the expertise itself becomes less valuable.

We believe that assumption is wrong: AI is not replacing expertise it’s lowering the knowledge requirements to be minimally competent on areas that until now were not accesible for most people.

For decades, professional value was built around access to knowledge. The people who knew the most were usually the people who created the most value. Engineers accumulated experience with specific technologies. Recruiters built an intuition after interviewing thousands of candidates. Managers learned to recognize patterns after years of leading teams. Knowledge was difficult to acquire, expensive to maintain and therefore highly valuable.

That world is changing remarkably fast. Today, almost anyone can ask an AI system to generate code, summarize a legal document, analyse financial information or explain a technical concept in seconds. Access to knowledge has become almost universal. What remains scarce is judgement.

The scarce resource is judgment.

Knowledge has become abundant. Criteria has not.

A senior engineer can produce significantly more code than before. A recruiter can evaluate many more candidates. A lawyer can process substantially more information. In every knowledge profession, AI increases the amount of work a single person can execute.

What AI does not do is is substitute critical thinking on critical tasks.

Someone still has to determine whether the code is maintainable, whether the candidate will thrive inside the team, whether the legal advice makes sense in context or whether the business problem being solved is even the right one.

In other words, AI dramatically reduces the cost of producing compelling answers. It’s actually the opposite, as the bottleneck is not producing anymore understanding became the critical skill, and sadly this is even more scarce

That distinction may become one of the defining characteristics of the next decade.

The gap between average and exceptional people will probably grow

One of the most interesting consequences of AI is that it simultaneously democratizes capability while increasing differentiation.

Everyone now has access to extraordinary tools. The barrier to entry is lower than it has ever been. Someone who could not write software a year ago can now build applications. Someone without design experience can create reasonably good interfaces.

That is undeniably positive.

At the same time, the best professionals are becoming dramatically more productive because they combine these tools with something AI still cannot provide: experience, context and judgment.

Technology amplifies everyone. But it also amplifies some people much more than others.

Throughout history, every major technological shift has followed this pattern: The Industrial Revolution reduced the importance of physical strength but increased the importance of people capable of organizing increasingly complex systems. Artificial intelligence is doing something similar with cognitive work: It is outsourcing part of our execution while increasing the value of the people capable of deciding where that execution should be directed.

This changes how companies should think about talent

Many hiring processes still assume that technical knowledge is the primary differentiator between candidates. Companies continue to evaluate programming languages, frameworks, certifications and technical exercises as if these were the strongest predictors of future performance.

They remain important, but they no longer tell the whole story.

When almost every candidate can produce competent code with the help of AI, organizations need to understand something much harder to evaluate. Can this person think independently? Do they understand trade-offs? Can they distinguish between an elegant solution and an effective one? Will they solve the customer's problem or simply produce more output?

These qualities have traditionally been described as soft skills, although there is very little that is soft about them. They are rapidly becoming the foundation of professional performance.

The irony is that as technical execution becomes easier to automate, evaluating people becomes significantly harder.

The same shift is happening inside engineering

Engineering has always struggled to measure contribution objectively. In the past, many organizations relied on activity because activity was relatively expensive.

Artificial intelligence breaks that assumption.

When thousands of lines of code can be generated in minutes, measuring activity becomes increasingly disconnected from measuring value. The interesting question is no longer how much someone produced. It is whether what they produced improved the product, accelerated delivery, reduced complexity or helped the business move forward.

The future belongs to people who exercise judgment

Identifying, developing and rewarding the people who consistently exercise good judgment in an environment where execution is increasingly automated is the true challenge when knowledge is no longer enough.

The real competitive advantage on the AI era is the ability to understand problems deeply, make better decisions and apply technology where it creates genuine impact.

About Dave García

Dave García is Co-founder of Pensero, the AI-era engineering performance and Agentic Deployment Intelligence platform.

Pensero brings together real signals from GitHub, Jira, AI coding tools, agents and the systems engineering teams already use to understand how work actually happens. By connecting delivery, quality, collaboration and AI usage, it helps leaders measure performance objectively, understand the ROI of their AI investments and guide their AI transformation, deciding where work is best performed by humans, augmented by AI or delegated to agents.

About Carles Font

Carles Font is Founder of Q-tech.

For more than 25 years, Q-tech has helped technology companies build exceptional engineering organizations through executive search, leadership advisory and talent strategy. Working alongside startups, scaleups and large enterprises, Q-tech helps organizations identify, attract and develop the people who will shape the next generation of technology.