The question AI cannot answer for you

Junior employees now have instant access to better answers than many senior people had five years ago. Ask an AI tool the right question and you get a framework, a recommendation, a draft. The problem Grover sees across the organizations NIIT works with is that easy access to answers has made the harder skill, knowing when to question them, even more important.

“The good performers treat AI as an accelerator for thinking and augmenting their problem-solving approach. They question the responses, validate the assumptions, and apply context and reasoning before making decisions. While the others risk accepting AI-generated responses at face value without fully understanding the underlying rationale.”

The capability gap this creates is real and largely invisible in most review processes. Two people can hand in the same AI-assisted deliverable, one having interrogated every step of it and the other having accepted the output without engaging with the reasoning behind it, and most managers will not spot the difference until something goes wrong.

“What we are witnessing is a shift in the nature of critical thinking itself. Success is no longer defined by who can find the answer fastest, but by who can ask the right questions, interpret insights effectively, and exercise sound reasoning and judgment. In an AI-enabled workplace, these are the capabilities that will define high-potential talent.”

What happens to learning when AI does the entry-level work

For decades, expertise in most professional roles built up through repetition. You did the routine work, you made the small mistakes, you developed judgment gradually. AI has absorbed much of that routine work, and with it, a significant portion of the learning that used to happen inside it.

Grover’s argument is that organizations cannot simply absorb this shift and carry on. The learning that used to happen organically now has to be designed deliberately.

“As AI takes over most of the routine and entry-level tasks, organizations must not reduce on the ask on how foundational skills of these new hires are developed. Traditionally, expertise was developed through repetition and gradual exposure. In an AI-first workplace, capability must be cultivated far more intentionally through meaningful experiences that develop understanding, judgment, and adaptability.”

In practice, that means structured mentorship, scenario-based learning, simulations, deliberate practice, and real-world projects. AI should sit inside these experiences as a coaching layer that accelerates feedback and reflection, giving people more reps at the thinking work even when the execution work is automated.

“The focus must shift from learning by doing repetitive tasks to learning by solving increasingly complex problems. AI can handle execution, but people must learn to ask the right questions, interpret context, challenge recommendations, make informed decisions, and apply sound business judgment.”

Where human judgment has to stay

Ask most people which decisions should stay with humans and you get the predictable list: hiring, strategy, anything ethically sensitive. Grover’s answer is more interesting. He does not think the right frame is keeping AI out of certain categories. The question is about who owns what comes out.

“I do not believe there are many areas that should be completely AI-free. In fact, AI should participate in almost every decision-making process by providing insights, identifying patterns, evaluating alternatives, and reducing cognitive load. However, there are certain decisions where the final accountability must always remain human.”

The decisions he has in mind are ones where the data cannot carry the full weight of the decision. Leadership selection, employee performance and career decisions, customer grievance resolution, regulatory compliance, strategic investments, decisions with significant societal consequences. What these share is that they require contextual understanding, values, and judgment that models cannot produce.

“Decisions involving people, ethics, trust, and long-term organizational direction require far more than analytical intelligence. These require contextual understanding, empathy, values, and sound judgment — qualities that cannot simply be inferred from data.”

On empathy, he makes a point worth noting. As more of the operational work gets automated, the ability to build trust, read a room, and take moral responsibility becomes a more distinctive leadership quality, not a softer one.

“AI can help us make better decisions, but it cannot build trust, inspire people, navigate ambiguity, or take moral responsibility. Those remain fundamentally human responsibilities.”

Measuring competence when AI is doing half the work

If AI can help almost anyone produce stronger output, faster, organizations face a real measurement problem. What are you actually evaluating when you review someone’s work?

“Organizations will need to rethink how they assess performance. In an AI-enabled workplace, output alone is no longer a reliable measure of capability because AI can help almost everyone produce content, analysis, and recommendations faster than ever before. The real differentiator is the quality of human thinking behind the output.”

Grover’s view is that assessment has to follow the same logic as learning redesign. The evaluation shifts to the thinking behind the work. Can someone frame the right problem? Can they identify when an AI recommendation is plausible but wrong? Can they make a judgment call when the data runs out?

“The focus should shift from what employees produce to how they think. Can they frame the right problem, challenge AI-generated recommendations, apply domain knowledge, navigate ambiguity, and make sound decisions? These are the capabilities that will define future success.”

What managers get wrong about efficiency gains

Automation frees up time. Most organizations are not being deliberate enough about where that time goes.

“Managers should encourage teams to use AI to eliminate repetitive work, freeing up time for higher-value activities such as innovation, strategic thinking, customer engagement, and cross-functional problem-solving. Time saved through automation should be reinvested in learning, growth, and finding new avenues to add value to work outcomes.”

The culture question matters here. Learning scheduled in quarterly bursts around operational pressure tends to lose to the operational pressure. Grover’s view is that it has to be embedded into the flow of work itself, built into how people actually spend their days.

“AI may improve efficiency, but long-term competitive advantage will still come from people. The organizations that thrive will be those that use AI to amplify human potential, not simply automate tasks.”