Physics Fridays — Paper No. 26

AI Isn't Losing. Complexity Is Winning.

Author: Robert Dvorak

Founder, BlueHour Technology

June 5, 2026

---

Throughout my career, I have had the opportunity to participate in hundreds of Digital Transformations, Business Transformations, cloud migrations, application modernizations, acquisitions, integrations, operating model redesigns, and large-scale technology initiatives.

The technologies changed.

The industries changed.

The business objectives changed.

What never changed was the underlying goal: create more business value.

Sometimes the value came in the form of revenue growth. Sometimes it came through cost optimization, improved customer experiences, risk reduction, operational efficiency, or competitive advantage. Regardless of the objective, every transformation ultimately came down to the same question: could the organization improve its ability to convert effort, investment, and intelligence into measurable business outcomes?

After four decades of watching organizations attempt to answer that question, I have become convinced of something that receives far less attention than it deserves.

The greatest obstacle to business performance is rarely technology.

It is complexity.

Complexity accumulates slowly. It arrives through acquisitions, reorganizations, technology investments, regulatory requirements, new products, new markets, and thousands of local decisions that made perfect sense when they were made. Over time, however, complexity becomes a force of its own. It begins slowing decisions, obscuring accountability, increasing costs, weakening governance, and making risk more difficult to identify and manage.

Most organizations don't recognize the problem until they begin feeling its effects.

Execution slows.

Priorities become less clear.

Change becomes harder.

Costs rise faster than expected.

Technology investments produce less value than anticipated.

The organization remains busy, but somehow becomes less effective.

This observation matters because we are currently living through one of the

most important technology cycles in modern history.

Artificial Intelligence is advancing at an extraordinary pace. New models, agents, platforms, and capabilities are appearing almost weekly. The amount of intelligence available to enterprises today would have seemed unimaginable only a few years ago.

Yet despite this remarkable progress, a question continues to surface in boardrooms and executive meetings across the world.

Where is the business value?

I do not believe the answer lies in the quality of the models.

Nor do I believe the answer lies in a lack of investment, ambition, or executive

commitment.

I believe the answer lies in the operating model itself.

Most organizations are attempting to deploy twenty-first century intelligence inside operating models that were designed for a very different era. As a result, the additional intelligence often encounters the same friction, handoffs, governance gaps, process inefficiencies, organizational silos, and decision bottlenecks that existed before AI arrived.

More intelligence enters the system.

The value creation curve does not move proportionally.

Which is why I have come to believe that AI isn't losing.

Complexity is winning.

Complexity becomes particularly dangerous when it begins growing faster than an organization's ability to manage it. At that point, complexity stops being a byproduct of success and becomes a constraint on future performance.

I often think about this through the lens of physics.

In physical systems, entropy represents the gradual loss of usable energy. Enterprises experience something remarkably similar. As complexity increases, organizations begin consuming more energy coordinating work than creating value. More effort is required to produce the same outcome. More meetings, approvals, handoffs, workarounds, and layers of oversight emerge. The organization remains active, but the amount of usable economic energy available to create value begins to decline.

Runaway complexity accelerates entropy.

Entropy eventually creates what we call Entropic Outages.

Unlike traditional outages, Entropic Outages are not immediately visible. Systems remain online. Employees continue working. Customers continue interacting with the business. Yet decision quality deteriorates, governance weakens, execution slows, and risk becomes increasingly difficult to identify. Over time, the enterprise begins consuming more energy than it creates.

Left unchecked, these conditions create Butterfly Effects, where seemingly minor issues cascade through interconnected systems and create disproportionate business consequences. In extreme cases, they can lead to what we describe as Humpty Dumpty Outages—situations where restoring operational effectiveness becomes dramatically more difficult than preventing the breakdown in the first place.

This is one of the reasons I believe the discussion around operating model modernization needs to evolve.

When executives hear phrases such as "rewire the operating model," they often imagine a massive, multi-year transformation effort with significant risk, cost, and disruption. Those concerns are understandable.

The reality is that enterprises are not monolithic operating models.

They are collections of operating models.

At BlueHour, we believe every enterprise operating model is actually the composite of dozens of domain-specific operating models, each responsible for delivering a specific business outcome. We refer to these as Micro Operating Model Units, or MOMUs.

Customer onboarding is a MOMU.

A claims process is a MOMU.

A sales motion is a MOMU.

A finance workflow is a MOMU.

A service delivery capability is a MOMU.

Each has its own stakeholders, technologies, governance requirements, performance metrics, and value streams. Each can be independently measured, modernized, and optimized. Most importantly, each can deliver measurable business value and measurable return on investment.

This changes the modernization equation completely.

Organizations no longer need to place a single large bet on transforming the entire enterprise. They can modernize one MOMU at a time, producing measurable outcomes, reducing complexity, and building confidence through a series of controlled successes. Over time, those successes compound into an Enterprise Portfolio of Modern Operating Models that systematically replaces the Traditional Operating Model without disrupting the business.

This is where BlueHour's Operating Architecture becomes important.

Much of the market is focused on intelligence itself. Models are becoming smarter. Agents are becoming more capable. Platforms continue to expand their functionality.

We view the challenge differently.

The most important question is not how much intelligence exists.

The most important question is whether the enterprise can convert

intelligence into outcomes.

That perspective led us to a simple framework:

AI × IT × Human Intelligence × Operating Architecture = Economic Energy.

AI contributes capability.

IT provides infrastructure.

Human Intelligence contributes judgment, creativity, experience, leadership, and accountability.

Operating Architecture determines whether those forces work together constructively or destructively.

When aligned, they generate Economic Energy—the ability to convert intelligence into revenue growth, cost optimization, risk reduction, stronger governance, improved customer experiences, and superior business performance.

This is the foundation of what we call Extreme Operating Leverage.

One of the more interesting developments in the AI market is that many leading AI companies are now investing heavily in deployment and professional services organizations. I view this as a healthy and important signal. It suggests the market is beginning to recognize that deployment and value realization are not the same thing.

Deployment introduces intelligence into the enterprise.

Operating Architecture enables the enterprise to convert that intelligence into

outcomes.

The challenge is deeper than implementation.

The challenge is operationalization.

And operationalization is fundamentally an operating model challenge.

The organizations that ultimately lead the AI era will not necessarily be those with access to the most advanced models. They will be the organizations that learn how to systematically reduce complexity while increasing intelligence. They will understand how to modernize operating models without disrupting the business. They will know how to generate more Economic Energy while reducing entropy.

Most importantly, they will recognize that AI is not the destination.

It is the opportunity.

The opportunity to redesign how intelligence, technology, and people work together.

The opportunity to achieve Extreme Operating Leverage.

The opportunity to build organizations that are more productive, more resilient, more transparent, more accountable, and ultimately more human.

If your organization is struggling to move from AI experimentation to AI value realization, I would encourage you to stop asking which model you should deploy next.

Ask a different question.

Which Micro Operating Model Unit should we modernize first?

The answer to that question may reveal where your next wave of business value is waiting to be unlocked.

Serving Business. Serving Humanity. Serving Truth.

Designed with Physics. Deployed with Economics. Determined by Humans.

— Robert Dvorak

Physics Fridays

---

Physics Friday - Paper No. 26.pdf

Download PDF • 245KB