What Liminal Leadership Actually Looks Like
The leaders navigating this moment best aren’t the ones with perfect forecasts. They’re the ones willing to operate in both worlds at once; managing today’s operations while building tomorrow’s capabilities. Here’s what that looks like in practice.
At Alder Hey Children’s Hospital, they built “the Batcave.” When we visited, Ian Hennessey showed us something remarkable: a 39-person innovation hub inside one of the UK’s leading pediatric hospitals, run by a practicing surgeon who splits his time between saving lives in the operating theatre and prototyping the future in the lab.
The team applies robotics, rapid prototyping, and AI to improve both operational efficiency and life-saving surgeries, all within the highly regulated NHS environment. They’re not choosing between current care and future innovation. They’re doing both, simultaneously. They exist on the threshold between life-critical operations today and a high-tech future tomorrow, ensuring innovation stays grounded in frontline reality.
That’s dual-state management. Not transitioning from one model to another, but holding both at once.
NVIDIA adopted “Pilot in Command” decision-making. Jensen Huang told us about their shift away from traditional hierarchy toward what they call “Pilot in Command”, similar to Netflix’s “Informed Captain” model. Power moves to the most informed person, not the most senior. The person closest to the decision makes it.
The result? NVIDIA now moves at what Huang calls “nanoseconds rather than years,” becoming one of the most valuable companies in the world by treating speed and informed action as more valuable than executive approval.
In a liminal moment, when the landscape is shifting faster than any single leader can track, the leadership role changes. You’re no longer the bottleneck of approval. You’re the facilitator of informed action. That’s a fundamentally different job.
At a recent event we hosted, we heard from Sofia Lim Oliver from HiBob how they turned AI adoption into a culture shift, not a tech rollout. Sofia, Director of Director of Leadership Development at HiBob, shared how an “AI Mind” initiative helped employees identify their biggest time-wasters and solve them using AI. Not top-down deployment. Bottom-up experimentation.
We subsequently explored, how navigating that “divergent phase”, of high experimentation with lots of learning, can result in re-imagined end-to-end processes. The key? To treat AI as a mindset and culture shift, not purely technological. By building confidence and capability first, organisations can bridge the threshold between human resistance and technological adoption.
What ties these examples together? None of these leaders waited for a clear playbook. They recognised the threshold, named it, and started building for what’s coming while still delivering on what’s here.
What Poor Liminal Leadership Looks Like
But let’s be honest about what happens when leaders get this wrong.
We’ve watched professionals with 25 years of expertise respond to AI by going silent. They view their intellectual property as a fortress to guard rather than a resource to evolve. They refuse to feed their knowledge into new systems because they’re petrified it will make them redundant.
It’s a very human fear. But it’s also a trap. As Sam Altman put it, they end up “focusing entirely on making the irrelevant more efficient.”
The consequences are real. While they spend time in echo chambers the world changes at the speed of light around them. If your expertise can’t be “reasoned about” by AI, you simply cease to be surfaced to the people who need you.
This is the opposite of liminal leadership. Instead of crossing the threshold, they’re trying to barricade it. Liminal leaders hold ambiguity without forcing premature resolution. They invest in long-term transformation while managing present operations. They build culture and relationships as primary strategy, not afterthoughts.
Leaders stuck in the old model optimise harder, retreat into protection, and wake up one day to discover the future chose someone else.
Why Kindness Is Your Operating System
Here’s where most leadership writing goes soft. But stick with us, because the evidence is hard.
Google’s Project Aristotle studied hundreds of teams to understand what made some effective and others not. The single biggest predictor wasn’t expertise or seniority. It was psychological safety, people feeling safe to speak up, make mistakes, and be themselves.
High-psychological-safety teams showed more innovation, better decisions, higher engagement, and lower turnover. A 2024 study on empathetic leadership found that empathy from leaders increased employees’ career adaptability and innovative behaviour, especially for people who naturally dislike uncertainty. A 2025 meta-analysis concluded that supportive, relational leadership has a significant positive effect on organisational adaptability.
Translation: kind, empathetic leadership isn’t fluffy. It’s the precondition for people to speak up, experiment, and adapt when no one has the full answer.
In a liminal moment, lack of kindness isn’t neutral. It actively destroys the trust and energy you need to move through uncertainty. When leaders respond to volatility with silence, denial, or coldness, employees report higher stress, less willingness to go the extra mile, and higher intention to leave.
But when leaders show up with both competence and care? When they acknowledge difficulty, treat people with dignity, model vulnerability? That’s when organisations hold together through the threshold.
This is what Sarah Gillard means when she says “Please, be kind. Especially when we don’t know what’s going on.” She’s describing the infrastructure you need to adapt. If you want people to experiment and stay while the ground is moving, psychological safety and kind leadership aren’t perks. They’re the operating system.
