Recruiting in the Age of AI - What You Should Expect from Your Engineers and Developers

As AI tools become widespread in companies, the skills traditionally sought in engineers and developers need to be reconsidered. Knowing how to code or perform a technical analysis is no longer sufficient. You must adapt your expectations and identify new forms of professional intelligence.

AI as a Copilot: The End of the Cult of Raw Skill

The emergence of AI assistants, coding copilots, and autonomous agents is transforming task distribution. Today, a developer can produce a functional prototype in just a few hours with the help of ChatGPT or Claude, where it used to take several days. It's no longer about doing everything yourself but knowing how to leverage AI to improve efficiency.

What to Really Look For: The Ability to Ask the Right Questions

The key skill is no longer technical but cognitive: knowing how to formulate a problem, describe an expected outcome, and evaluate the relevance of a response generated by an AI agent. As an expert expressed during the INSA roundtable, "all the responsibility lies in the question." Asking a good question to an AI means being able to express the quality criteria for the expected result.

For example, asking "Generate an auditorium plan" is not enough. A good engineer will specify: "I want every person to have a good view of the stage, optimal acoustic comfort, and a system that is adaptable for different types of events." This ability to define a set of criteria—what constitutes a good solution—is now the cornerstone of engineering work.

Another example, in software development: a developer will not be judged on their ability to code a function but on how they precisely document what it should accomplish, why, and in what context. The AI handles execution; the human remains the strategist of meaning.

Key notions are emerging:

• Technical curiosity: to explore different angles before making a request.
• Critical thinking: to compare AI responses with real-world constraints.
• Clarity of expression: to write effective and contextualized prompts.
• Abstraction skills: to translate an operational need into computational logic.
• Ethics and vigilance: to question biases and the consequences of a generated result.

A high-performing engineer is thus a designer of relevant questions, a modeler of intentions. They know that the quality of the output entirely depends on the quality of the input. They don't just speak the language of code but the language of reasoning.

A Good Mediator Is Better Than an Isolated Expert

In a world where AI agents become hyper-competent specialists in very specific domains, human value lies in the ability to orchestrate them. A profile capable of making different agents collaborate, transmitting the right parameters, and validating results will bring you much more than a solitary coder. It's a matter of orchestration, not accumulation of knowledge.

A good candidate will tell you: "I don't know this tool yet, but here's how I would approach it to get the best out of it, and here are the questions I would ask to evaluate its use." The ability to learn autonomously has become a more reliable recruitment criterion than static expertise.

Recruitment Should No Longer Be Based on Mastered Tools

Competence in a specific language, framework, or AI tool has become perishable. What matters is the ability to learn quickly, adapt, and maintain a critical mindset toward automated solutions. The best hires are those with the right mindset, not those with an obsolete certification.

Conclusion: Evolve Your HR Perspective

You are no longer recruiting just an executor or a technical expert. You are recruiting someone capable of dialoguing with AI, clearly formulating intentions, and ensuring the impact and coherence of the work produced. The ability to ask the right questions, to arbitrate between generated results and business requirements, is becoming a strategic skill. What you're looking for is not a "complete" profile in the traditional sense but a "combining" profile: someone capable of making humans and machines collaborate in a logic of augmented performance.

In HR, this means redefining job descriptions, revising evaluation grids, and including reasoning and problem-formulation tests in interviews. You're no longer looking for technical know-how but situational intelligence, tool awareness, and a conductor's mindset. This perspective will make all the difference.