top of page
Search

AI Developers Are Training Their Neural Networks. Why Aren’t We Training Ours?


AI companies are investing tens of billions to train their neural networks—continuously, deliberately, and at scale. This illuminates a basic truth: performance does not emerge from static knowledge alone; it emerges from trained systems.


Humans, by contrast, are performing in increasingly complex, fast-moving, and emotionally demanding environments—yet we’re investing almost nothing financially or structurally to retrain our own neural architecture.


As the value of personal knowledge is being eclipsed by generative AI, this asymmetry is becoming stark—and increasingly critical. The result is a growing gap between what humans know and how reliably they can perform under pressure.


Hidden Assumptions About Human Capabilities


Beneath this imbalance lies a familiar and largely unexamined assumption: that human capability naturally scales with growing information and effort.


We assume that knowledge leads to competence, that pressure builds resilience, and that understanding how something should be done eventually translates into the ability to do it well.


This assumption is embedded in our legacy education, leadership development, and organizational training. It shapes how we teach, how we manage, and how we respond when performance falls short.


Neuroscience challenges this assumption in a fundamental way.


Knowledge Under Pressure


Human performance is not determined simply by what we know or how motivated we are. It’s fundamentally impacted by early neural systems that govern attention, emotional regulation, and threat detection—a finding supported by decades of research in cognitive neuroscience and stress physiology.


These systems operate largely outside conscious awareness and are hard-coded into primitive regions of the brain. Under pressure, they interrupt the prefrontal cortex. Attention narrows. Working memory collapses. Previously learned knowledge remains stored, but it is less accessible when needed most.


This disintermediation of our faculties has become commonplace, as overwhelming stress, distraction, uncertainty, conflict, and burnout continue to accelerate at home and in the workplace. At the very moment we most need our intelligence and expertise, our biology suppresses them.


Interrupting Ancient Circuits 


For more than a century, medical doctrine held that the adult brain was static, asserting that a stable population of neurons was a biological necessity for maintaining long-term memories and learned behaviors. But modern neuroscience has upended that convention—decisively confirming that neurons continue to form through neurogenesis and physically change through neuroplasticity.


Today, we know that the mind can change the brain—and that its network instructions can be intentionally altered through specific neural practices. In fact, it turns out that many human capabilities, once thought to be immutable traits, are in fact trainable mental skills.


This means we have the capacity to intentionally change our default responses to the immediate, automatic arousal signals of stress, the habitual interpretation of social challenges as mortal threats, and our attachment to limiting beliefs and biases. These, along with other mental and emotional skills, have become critically important for leaders and professionals in the workplace.


The Urgency of Human Skills


As AI accelerates decision cycles and VUCA conditions become the permanent operating environment, the limiting factor in organizational performance is no longer technical expertise or access to information. It is the reliability of human mental and emotional skills under pressure. Attention control, self-awareness, emotional regulation, adaptability, and interpersonal effectiveness are becoming core competencies for leadership and workforce execution.


Yet these human skills are precisely those least supported by our traditional education and skills-training systems, which are built on an explicit-learning framework.


As a result, we can learn about empathy without becoming more empathetic—understand what authenticity is without feeling any more authentic—or know that anger is not an appropriate response, but have no way to anticipate or interrupt it.


Human skills are governed by neural systems that operate independently of explicit knowledge. When these systems are untrained, they default to stress-driven responses regardless of experience, intelligence, or intent—and habitual attachments to non-conscious beliefs and biases


Our Human Skills Gap


We're experiencing a widening gap between the skills modern work demands and the methods we use to develop them. Closing that gap requires a shift from conventional explicit learning to procedural training—deliberate practices that create embodied, experiential conditions that can intentionally shape or retrain neural networks.


Bringing procedural methods into our educational, leadership development, and professional training frameworks is urgent to ensure effective human performance amid increasing pressures.


Training Our Own Neural Networks


The idea of training the mind is not new. For millennia, contemplative traditions such as Buddhism and Stoicism developed disciplined practices to cultivate attention, emotional regulation, and equanimity. In the twentieth century, modern psychology extended this work through evidence-based cognitive behavioral therapy and positive psychology, translating neural training into clinical and behavioral contexts.


What is new is the convergence of these traditions with contemporary neuroscience and artificial intelligence. Decades of research into how the brain learns, adapts, and reorganizes—alongside parallel advances in training large-scale artificial neural networks—have clarified many underlying mechanisms that make training effective. We now understand far more precisely how attention, emotion, perception, and response patterns are conditioned, stabilized, and changed over time.


Crucially, many of these insights are translatable. Practices that once relied on tradition or intuition can now be designed, explained, and applied with scientific rigor. This makes it possible to bring neural training out of monasteries and clinics and into mainstream education, leadership development, and workforce training—without mysticism and without guesswork.


Seen this way, training our own neural networks is not a metaphor. It is a practical extension of what we are already doing with artificial systems: identifying the skills required, designing conditions that reliably train them, and reinforcing those capabilities through deliberate practice.


As AI and VUCA change accelerates, the brain increasingly constrains performance—and so procedural retraining of our own neural networks should be an essential new focus for us and an important new element of education and organizational learning.



References


McEwen, B. S., & Sapolsky, R. M. (1995–2017). Stress, allostasis, and allostatic load: Implications for neurobiology and health. Annals of the New York Academy of Sciences; Physiology & Behavior.


Arnsten, A. F. T. (2009–2015). Stress signalling pathways that impair prefrontal cortex structure and function. Nature Reviews Neuroscience.


Squire, L. R., & Dede, A. J. O. (2015). Conscious and unconscious memory systems. Cold Spring Harbor Perspectives in Biology.


Schwabe, L., & Wolf, O. T. (2013). Stress and multiple memory systems: From “thinking” to “doing.” Trends in Cognitive Sciences.


Posner, M. I., & Rothbart, M. K. (2007–2018). Research on attention networks and self-regulation. Developmental Psychology; Proceedings of the National Academy of Sciences.


Tang, Y.-Y., Hölzel, B. K., & Posner, M. I. (2015). The neuroscience of mindfulness meditation. Nature Reviews Neuroscience.


Merzenich, M. M. (2013). Soft-wired: How the new science of brain plasticity can change your life. PLOS Biology (and related research).


Davidson, R. J., & McEwen, B. S. (2012). Social influences on neuroplasticity. Nature Neuroscience.


Aboumoussa, L. (2024). Leadership Development in the Age of Artificial Intelligence. Harvard Kennedy School.


International Coaching Federation (ICF). Global Coaching Industry Research 2025.


McKinsey & Company (2023–2025). Generative AI and the Future of HR; AI in the People Function.


The Conference Board (2025). AI-Powered Coaching and the Future of Career Development.


Terblanche, N. H. D. (2024).


Artificial Intelligence (AI) Coaching: Redefining People Development.” Journal of Applied Behavioral Science. (2025)


World Economic Forum (2025). Future of Jobs Report 2025.



Organization: Institute for Organizational Science and Mindfulness (IOSM)






About IOSM


The Institute for Organizational Science and Mindfulness (IOSM) is a global association of human capital and operating leaders, educators, and coaches. We share a common mission to apply neuroscience and neural training to develop more effective leaders, a happier, healthier, and higher-performing workforce, and a safer, more inclusive, and more productive workplace.

 
 
 

Comments

bottom of page