Why Productivity Stalled Long Before AI Arrived, And What Leaders Are Still Missing
- Jan 20
- 6 min read
Written by David Bovis, Founder of Duxinaroe Ltd.
David Bovis is a leadership strategist and founder of Duxinaroe, specialising in the neuroscience of decision-making, behaviour, and performance. Creator of the BTFA (Believe-Think-Feel-Act) framework, he works with senior leaders to address the neurological root causes of misalignment, disengagement, and failed change.
The last few years have been framed as a turning point for productivity. Artificial intelligence, machine learning, automation, and advanced analytics are widely described as the long-awaited cure for stagnating performance. Boards are told that smarter systems will finally unlock efficiency, insight, and growth at scale.
And yet, the uncomfortable truth is this: productivity stalled long before AI arrived.
Across the UK, long-term data from the Office for National Statistics and the National Institute of Economic and Social Research show that productivity growth began slowing decades ago, well before the digital acceleration of recent years. A similar picture appears across other advanced economies. Despite successive waves of technology, output per hour has never returned to the post-war trajectory leaders often assume to be the natural state of progress.
At the same time, Gallup’s State of the Global Workplace continues to report that close to four-fifths of employees worldwide are not engaged in their work. Misalignment, disengagement, and resistance to change have become enduring features of organisational life, not temporary side effects of disruption.
The paradox is hard to ignore. We have never invested more in systems, data, and control, yet performance improvement has flattened, and human energy has declined. AI did not cause this problem. It simply arrived into it.
The repeating promise of technology
This is not the first time leaders have been told that a technological leap would resolve deep-seated performance challenges.
Over the past four decades, organisations have moved from MRP to MRPII, ERP, RPA, machine learning, and now AI. Each wave arrived with the same promise: better information, tighter control, improved coordination, higher productivity.
Each wave delivered genuine benefits, but none reversed the underlying trend. The issue is not technology itself. The issue is the assumption that sits beneath it.
Most leadership and management systems have been built on the belief that performance improves when processes are optimised and variability is reduced. People, within this logic, are expected to comply, execute, and adapt to the system. When output falls short, the response is usually more structure, more measurement, or more automation.
What is rarely questioned is how those systems are experienced by the people expected to perform within them.
Where leadership models quietly fall short
Modern leadership education is strong on strategy, governance, targets, and incentives. It is far weaker on how human beings actually think, feel, and decide under pressure.
Leaders are trained to manage outputs, but not to understand how belief, perceived threat, or psychological load shape judgement and behaviour. The human brain is treated as a rational processing unit, rather than a biological system constantly balancing performance with survival.
This gap matters because performance does not emerge from systems alone. It emerges from brains interacting with systems, goals, and each other.
When environments are experienced as controlling, punitive, or misaligned, the brain shifts into protection mode. Attention narrows, learning slows, and problem-solving gives way to risk avoidance. Compliance may increase, but adaptability declines.
Seen through this lens, disengagement is not a motivation problem. It is a neurological response to context.
Evidence in practice: Different assumptions, different outcomes
The contrast between Toyota and Volkswagen is most visible when comparing production units per employee, a standard operational measure of how effectively human capability and production systems interact: Toyota produces roughly 10.8 million vehicles with approximately 384,000 employees, while Volkswagen’s global group output of about 9 million vehicles comes from around 680,000 employees, highlighting a significant difference in vehicles per capita performance. Reuters
Both organisations operate in the same global market, face similar regulatory pressures, and use advanced manufacturing technologies. Yet their approaches to people and problem-solving differ markedly.
Toyota’s system assumes that performance depends on developing people who can think, identify problems, and improve work at the point of activity. Standardisation exists, but it is designed to support learning, not replace it. Technology acts as an enabler of human judgement.
By contrast, more system-centric models rely heavily on control through process enforcement and centralised decision-making. As fewer people are trained to think and solve problems locally, more systems are required to compensate. The organisation becomes increasingly dependent on technology to manage complexity that the human system has not been allowed to absorb.
The result is not simply a cultural difference, but a neurological one. One environment reinforces agency, competence, and safety. The other unintentionally amplifies stress, dependency, and disengagement.
Psychological safety, without the biology
This is not an argument leaders have not encountered before. Over the past decade, Google’s Project Aristotle highlighted the importance of psychological safety in team performance. The concept was later popularised by Amy Edmondson and has since become a common feature of leadership conversations.
The intention was right. Teams perform better when people feel safe to speak up, challenge, and learn.
The limitation has been in how this idea is understood. Psychological safety is often discussed in subjective or behavioural terms, without grounding it in how the brain actually responds to perceived threat or support. As a result, it is treated as a cultural aspiration rather than a biological requirement.
Without that neurological understanding, leaders struggle to see how everyday decisions, targets, incentives, and systems either reinforce safety or undermine it. The language sounds progressive, but the operating conditions remain unchanged.
Longitudinal research into incivility, micro-management, and chronic stress has shown that the effects extend far beyond the workplace, influencing health, relationships, and even outcomes at home. These are not soft issues. They are indicators of how deeply organisational environments shape human functioning.
The missing layer: How performance really emerges
What most organisations are missing is not another tool, framework, or technology. It is an understanding of how performance emerges in the brain.
Under conditions of perceived safety, the brain allocates energy to learning, creativity, and problem-solving. Under conditions of threat, it prioritises protection, certainty, and short-term control. The shift is automatic, not chosen.
Technology-heavy environments that emphasise surveillance, metrics, and compliance often unintentionally signal threat, even when leaders believe they are driving clarity and efficiency. The result is a workforce that appears compliant, but is neurologically constrained.
This helps explain why engagement scores, innovation, and productivity plateau despite ever more sophisticated systems. Leaders are optimising the visible architecture of work, while neglecting the invisible biological one.
Where BTFA™ fits, and why it matters
This is the gap the BTFA (Believe-Think-Feel-Act) framework was designed to address.
Rather than starting with behaviour or culture, BTFA starts with belief. What leaders believe about people shapes how they design systems. Those systems shape how work is experienced. That experience drives emotional response, decision-making, and ultimately performance.
By making this chain visible, BTFA gives leaders a way to see why well-intentioned change efforts often fail, and why pressure amplifies the very behaviours they are trying to eliminate.
Importantly, BTFA is not a tool to impose behaviour. It is a form of leadership education that helps people recognise how their own assumptions interact with human neurobiology. Once that connection is made, many of the last thirty years of stalled improvement suddenly make sense.
A turning point, not a warning
We are not at the beginning of the AI era. We are at the beginning of a deeper reckoning with how leadership actually works.
The opportunity now is not to use technology to extract more from people, but to design systems that work with the human brain rather than against it. That requires a shift in belief, language, and education, from boardroom to shop floor.
Once leaders see that productivity, engagement, and adaptability are biological outcomes as much as operational ones, it becomes impossible to unsee. Performance stops being something that must be forced, and starts to be something that can be enabled.
AI may well accelerate this shift. But only if leaders first understand the system it is meant to support: the human brain.
Read more from David Bovis
David Bovis, Founder of Duxinaroe Ltd.
David Bovis is a leadership strategist and founder of Duxinaroe, specialising in the neuroscience of belief, decision-making, and performance under pressure. He is the creator of the BTFA (Believe-Think-Feel-Act) framework, a practical model that helps leaders understand why change, culture, and strategy often fail despite good intent. David works globally with senior leaders to address the neurological root causes of misalignment, disengagement, and stalled performance. His work bridges neuroscience, leadership, and systems thinking to enable sustainable behavioural change where traditional approaches fall short.
