I asked our panelists which blind spot concerned them most. Veloso zeroed in on workforce transformation, drawing a parallel to the Industrial Revolution: “We’re going to be replacing routine cognitive tasks with intelligent systems,” he said. Just as mechanisation reshaped work through Taylorism and assembly lines, AI will fundamentally change “the skill set, the organisation of firms, the fabric of what it means to be a worker”.
This isn’t just about individual companies adapting – it’s about reinventing how work itself functions.
Fang cut straight to the philosophical heart of the matter: “Before the how and what, it’s important to ask the why.” She compared this to how AI works through reinforcement learning, where researchers define the objective function for the machine to optimise. Organisations need similar clarity about what they’re trying to achieve with AI.
Weighing in on the issue of workplace transformation, Maciejko observed that there are two basic approaches: strategic top-down planning and change from the bottom-up. To him, the latter approach is more important yet often neglected by organisations.
“Find the top 7-10 percent of your team, the people who are already using AI, maybe secretly, and empower them fully.” Let them experiment, he said, find what works, then scale those solutions across your organisation.
When I opened the floor, our audience members didn’t hold back. Mary Falvey, President of Falvey Associates, warned that “we will soon reach the state where the rate of change in technology will outpace the rate of change in our ability to govern it”.
Rather than getting stuck in regulatory battles, she suggested focusing on principles – what behaviours do we want AI to encourage or discourage?
Amir Hosseini, Partner & Managing director at Alix Partners hit on something many executives miss: the unsexy but essential work of implementation.
“There is a huge gap in companies, what they need as a foundation in terms of data asset management, data strategy,” he noted. Having lived through waves of technology adoption, from function-specific applications to enterprise-wide ERPs (enterprise resource planning), Amir observed that AI requires a level of data readiness many organisations simply do not have.
Nicolas Vilmin, Managing Director and Executive Coach at Vilmin Coaching, raised a question that would resonate with most: “What is the true cost of implementing AI at the speed that we’re going?”
He wasn’t just talking about dollars but the human cost – stress, displacement and societal impact when technology adoption outpaces our ability to adapt.
The questions we’re asking too much
If what we’re not asking enough reveals what we’re missing, then the flip side is just as telling.
Our survey revealed a clear consensus: some questions are dominating the AI conversation – not because they’re the most useful, but because they’re the loudest. Leaders seem fixated on the technology hype and fear of missing out (FOMO), chasing trends, buzzwords and superficial innovation agendas, often tied to short-term gains and job replacement.
Fang noted that AI hype “creates a lot of busy action without clear direction” while feeding unnecessary fear. For executives feeling overwhelmed, Maciejko suggested a practical starting point: Develop “an AI radar of your own” by finding trusted sources – whether it’s on social media, podcasts or newsletters – and listening to people really in the know.
Veloso made a point that hit home for many: technology will continue advancing rapidly, but we need to focus less on tracking every new development and more on adapting our organisations to handle what’s already here.
“That’s what we do as a school. What we do through our intellectual leadership, both directly through the research but also by our extended network, is a good example of … helping people understand what’s working, what’s not working, what’s making a difference or not.”
Echoing this, Siddartha Chaturvedi, Director of Product Management at Microsoft, called out our collective “obsession with state-of-the-art” technologies. He challenged leaders to first ask whether the underlying processes they’re trying to automate should even exist in their current form. “Should these processes exist as is in the first place?” Often, the opportunity isn’t automating what we already do but rethinking it entirely.
Philip Brewer, Managing Director at Stoked Spoke, shared the question he repeatedly faced: “Can I replace you with AI?” He then posed a pointed challenge to AI-obsessed executives: If you get rid of everybody and just rely entirely on AI, and you’re asking it what to do, who are you actually listening to?
The tensions every leader faces
What fascinated me most were the patterns that emerged when we compared the under-asked and over-asked questions. Three key tensions stood out:
First, there’s what I call the paradox of repeated issues. Topics like workforce appeared in both categories, but with completely different framing. Fear-driven questions like “Will AI replace jobs?” are over-asked, while more proactive and forward-looking ones like “How do we build effective human-AI teams?” are under-asked. As Fang noted, this isn’t a contradiction but reveals the multi-faceted nature of these complex issues.
Second, we face a persistent priority mismatch between short-term and long-term goals. Veloso advised organisations to “do a little bit of both” – connect AI initiatives to your core strategic capabilities, while creating space for experimentation and serendipitous discoveries.
Finally, there’s the aspiration-implementation gap. Maciejko shared a striking statistic from a Boston Consulting Group survey of C-suite executives: 75 percent think AI is a top three priority, but only 25 percent see any significant impact from AI right now. This very same 25 percent are focusing on high-impact areas rather than scattered pilots, empowering their teams, establishing clear governance and developing strategic roadmaps that embed AI into workflows. They are playing the long game.
Where do we go from here?
Our audience raised thought-provoking questions about the future.
Morten Hansen, Visting Professor of Entrepreneurship and Family Enterprise at INSEAD and professor at the University of California, Berkeley asked whether AI poses an existential threat to business schools. Our panelists responded with confident optimism. Instead of diminishing the role of business schools, said Fang, the AI transformation actually “highlights the importance of business schools” in guiding leaders, organisations and society to adapt to the profound change.
Veloso emphasised the amazing opportunity for business schools to better reflect the reality of what we’re seeing in business. He described how the professor’s role will evolve towards being “a curator, a focal point, a coach to the learning journey,” incorporating real-world AI tools into classrooms to prepare students for the new business landscape.
Maciejko closed with a galvanising observation about INSEAD’s unique position.
“By sheer luck, we’re sitting here a hop from OpenAI in San Francisco. We have a campus in Abu Dhabi, the leader in Middle East AI. France is an AI leader in the EU. Singapore is one of the most future-forward countries on AI.”
He added to audience applause: “Nobody has these assets. The opportunity is enormous.”
As we wrapped up, I was struck by how profoundly this conversation extended beyond technology. In our rush to adopt AI, it is easy to chase tools and tactics – but the questions we choose to prioritise will shape not only the future of our businesses, but also the kind of society we create.
At INSEAD, we’re committed to helping leaders ask the questions that matter most in this transformative moment, because the future won’t be built by those with all the answers, but by those brave enough to ask the right ones. That might be the most important skill of all.
“INSEAD, a contraction of “Institut Européen d’Administration des Affaires” is a non-profit graduate-only business school that maintains campuses in Europe, Asia, the Middle East, and North America.”
Please visit the firm link to site
