When AI has all the answers, leaders need better questions
A conversation with Melanie Steinbach, EZRA advisory board member and former Chief People Officer at McDonald's and MasterClass, on curiosity, discernment, and leading when no one has the full picture.
Mel Steinbach has spent nearly three decades helping organizations solve complex people challenges. She brought EZRA into McDonald's — making it the company's first client — and now serves on the advisory board. Over her career she has built deep experience. And with it, strong opinions.
One in particular stands out: most organizations are asking the wrong question about AI.
"The magic happens with the second question," she says. "Not just 'Can you tell me X' — but once you get the answer, really starting to interact with it."
Speed of adoption, she argues, is the wrong metric. What will separate organizations over the coming years is the quality of human judgment being applied alongside the tools.
Curiosity is a leadership practice
At McDonald's, Mel found the organization had grown insular — optimizing internally rather than genuinely asking what customers needed next. The fix came from treating curiosity as a leadership behavior: named, modeled, and measured. Not a value on a wall, but something leaders were expected to demonstrate and be held accountable to.
In an AI context, the same logic holds.
Most people ask one question, get an answer, and stop. The leaders getting the most out of AI keep going — probing, pushing back, re-asking. Mel draws the comparison to learning an instrument. "We think of playful as just fun-loving. But you play a sport, you play an instrument — that requires practice, dedication, drills. That level of dedication is what we need around curiosity."
She calls it “playing with purpose.”
The four human differentiators in an AI powered workplace
EZRA's research points to four human qualities that will matter most as AI becomes table stakes:
Curiosity — asking better questions, not just more questions
Discernment — judging quality, risk, and relevance in context
Humility — staying a learner when certainty is impossible
Connection — building trust and alignment when work gets automated
When Mel first saw the framework, she says she "breathed a little sigh of relief that there's still going to be a need for people and culture teams for the foreseeable future."
The point she keeps coming back to is connection. As AI takes on more of the work that once brought teams together organically — the shared project, the collaborative document, the problem worked through together — connection becomes something leaders have to purposefully build. "The best leaders are really going to amplify that skill," she says. "My job is to get information from my team but more importantly to give it back. How quickly can I get that transferred to them?"
On discernment, she offers an insight that helps reframe the debate around generational use of AI. Younger employees tend to be more experimental and AI-native, while senior leaders bring deep pattern recognition for what good output looks like, even if they're less comfortable with trial and error. Neither group has the full picture. Both benefit from developing what the other already has.
And on humility, she's straight to the point. "Leaders are better when they are learners rather than knowers. If anyone says they know definitively right now across so many areas of AI — they are deluding themselves." The distinction Satya Nadella made at Microsoft — not know-it-alls, but learn-it-alls — captures the version of leadership that survives this moment. The executives Mel has observed sustaining real performance are the ones who have genuinely internalized that mindset.
How coaching helps teams lead through AI change
When a CEO says "we're all in on AI," Mel has seen what happens in the layers below. The direct reports understand it reasonably well. Three levels down, the message has gone fuzzy. By the time it reaches managers, fear has filled the gap that context left open. "Change does not move quickly in an environment of fear," she says.
Coaching bridges the distance between organizational direction and individual understanding, giving people space to ask the questions they can't ask publicly, work through what change means for them specifically, and show up differently as a result.
Her advice for leaders trying to move their organizations right now is stop pretending you know the answer, and start learning in public instead. "When you stop pretending, your trust levels with your team are going to soar. They believe you are lying to them when you tell them you know all the answers — and that's not serving anyone."
What she’s watching most closely isn’t adoption speed, but absorption — how quickly knowledge actually lands, gets processed, and changes behavior. "We are pedalling one side very quickly,” she says. “The knowledge absorption, processing, and application is likely to happen at a slower pace. As leaders, we're going to have to manage that gap,"
Listen to the full episode of EZRA Asks with Melanie Steinbach.