Eclectic Intelligence: Five Learnings About AI in Operations

ECLECTIC LEADERSHIP MOVEMENT | OPERATIONS & LEADERSHIP

No one tells you this about AI in Operations: the problem was never the technology. It was always the framework.

Every organisation I walked into over the last ten months had already bought the tools. Most had run the pilots. Several had given someone a shiny new title with "AI" somewhere in it. What none of them had was a coherent operational framework that told them where AI actually sat, what it owned, what it complemented, and what it most definitely should not be left alone to decide.

That gap is expensive. Gartner predicts that over 40% of agentic AI projects will be cancelled by the end of 2027, due to escalating costs, unclear business value, or inadequate risk controls. (Gartner, June 2025)

That is not a technology failure. That is a framework failure. And it is exactly the kind of problem the Troika of Operational Excellence was built to prevent.

Before we get into the detail, here is what this article will give you:

• The Troika of Operational Excellence, the first Eclectic Leadership framework built ten years ago, still holds. The outer ring has evolved.

• AI Agents are not tools. They are participants. Your People arm now includes Agents, and your OpEx budget needs to reflect that.

• Operational efficiency now has a third metric: tokens. Saving time and money is no longer sufficient.

• Corporate context matters more than corporate culture. You cannot write the room until you have read it.

• Language is the first leadership skill, and always has been. Large Language Models just proved the point.

• Eclectic Intelligence is a meta-capacity, not a job title. And it may be the most defensible skill set for the decade ahead.

If you prefer to listen on the go, then click the podcast.

The Framework That Held

Ten years ago I built a triangle. Three corners: Project Management, Operations, Delivery. Three arms: Process, Technology, People. Inside, four stacked layers from base to apex: Customer Specific Services, Commercials, Culture, with People running as the foundation beneath all of it.

I called it the Troika of Operational Excellence. The insight behind it was simple: the systems problem and the people problem are almost always the same problem. You do not fix one without the other.

That holds. After four organisations, two continents, and more governance reviews than I care to count, the core architecture has not broken. What it needed was an updated outer ring, and two adjustments inside.

The internal stack now reads: Customer Specific Services, Commercials, Context, Culture. Context sits immediately below Culture. That sequence is deliberate. You cannot shape a culture you have not first accurately read. Context is the diagnosis; culture is the intervention.

Context is the reading. Culture is the writing. Get the order wrong and you will spend years editing a draft you have never actually read.

The People arm now reads: People and Agents. And the outer ring of the triangle carries three new labels, one on each arm: Artificial Intelligence, Human Intelligence, and Eclectic Intelligence. That outer ring is not branding. It is the operating environment every Operations leader is now working inside, whether they have named it or not.

Learning One: Agents Are Participants, Not Tools

Gartner predicts that 40% of enterprise applications will be integrated with task-specific AI agents by the end of 2026, up from less than 5% in 2025. (Gartner, August 2025) That is not a technology adoption curve. That is a workforce composition shift.

AI Agents complete tasks, make decisions within parameters, trigger downstream workflows, and in some configurations manage other agents. Treating them as sophisticated automation is how organisations end up with agent deployments that nobody governs and everybody blames when something goes wrong.

Gartner notes that conventional ROI calculations do not fully reflect the unique cost and value dynamics of AI-agent-powered workflows. Which brings me to the third metric.

Learning Two: OpEx Now Has Three Metrics

Time. Money. Tokens.

Every unnecessary API call, every bloated prompt, every poorly scoped agent loop running in the background has a direct cost. Most Operations leaders are not yet tracking this. They will be. Token economics is not a developer problem; it is a budget management problem, and it belongs on the Operations dashboard.

McKinsey's State of AI in 2025 found that only 39% of organisations report any EBIT impact attributable to AI, despite 88% now using AI in at least one business function. (McKinsey, November 2025) The majority are spending freely and measuring loosely. Adding token efficiency as a formal OpEx metric is one of the fastest ways to close the gap between AI activity and AI value.

Learning Three: Context Before Culture

Here is the finding I want to argue for most plainly.

Operations is the primary culture carrier in any organisation. Not HR.

HR designs culture. Operations lives it, daily, in every process decision, every escalation, every vendor call, every moment where someone has to choose between the stated policy and the actual way things get done around here. Operations touches every other team. It is the veins of the body, carrying what is needed to every organ, every day. If the Operations team does not embody the culture, no values workshop will save you.

But before you can carry a culture, you have to read the context accurately. What is the real power structure here? What does this organisation actually reward, regardless of what the strategy deck says? Who are the informal connectors? Where does information pool and where does it flow?

Gartner identifies context as one of the most critical differentiators for successful agent deployments, noting that AI agents can interpret industry-specific context to make sound decisions even in unfamiliar scenarios. (Gartner, October 2025) What is true for agents is equally true for the humans managing them.

Cross-cultural training needs to graduate into cross-contextual training. The question is no longer only how people from different countries prefer to communicate. The question is what is the actual context this team is operating in right now, and what kind of intelligence does navigating it require.

Learning Four: Language Is the Original Leadership Technology

The first foundational discipline of Eclectic Leadership is Linguistics. That was not a whimsical choice ten years ago. It turns out to have been quietly prophetic.

A Large Language Model is, at its most precise, a statistical instrument trained on the patterns of human language. It finds regularities, predicts sequences, and constructs responses based on prior probability. Prompt engineering, the skill of directing an LLM toward useful output, is fundamentally an exercise in linguistic precision. The clearer your language, the better the result. The more ambiguous your instruction, the more the model fills the gaps with its own assumptions.

Sound familiar? That is leadership. Every briefing you have ever given. Every strategy document that confused more than it clarified. Every performance conversation that went sideways because the framing was wrong. Language awareness has always been leadership awareness. LLMs did not create that relationship; they simply made it impossible to ignore.

McKinsey highlights that organisations building AI familiarity across functions have accelerated capabilities in prompt engineering and model evaluation as core competencies. (McKinsey State of AI, 2025) Prompt engineering is not a technical skill. It is a linguistic one. And for any Operations leader managing a hybrid human-agent workforce, it is the new baseline.

Learning Five : Eclectic Intelligence, a Definition Worth Fighting For

Let me be precise about what it means, and what it does not.

Eclectic Intelligence is not Emotional Intelligence, though EQ is part of it. It is not Ethical Intelligence, though that belongs within it. It is not the fashionable union of critical thinking, creative thinking, and systems thinking, though all three are present. Those framings are useful. They are also incomplete, because they remain predominantly human-centric and static in a world that now requires fluency across human and artificial systems simultaneously.

Eclectic Intelligence is the meta-capacity to move intelligently between Artificial Intelligence, Human Intelligence, and the full spectrum of human wisdom traditions, drawing on each according to what the situation demands, rather than defaulting to whichever one you were trained in.

AI is extraordinary at pattern recognition, language modelling, context compression, and working at scale with consistency. It is, honestly stated, a very sophisticated statistical instrument that has absorbed most of what humans have ever written. What it lacks is moral judgement, the capacity to hold genuine ambiguity, and the wisdom that comes from having skin in the game.

Human Intelligence fills those gaps. But it carries its own limitations: cognitive bias, finite memory, emotional interference, and a persistent tendency to mistake familiarity for understanding.

The oral traditions of indigenous leadership frameworks understood something about memory and knowledge that we are only now rediscovering in the age of context windows and token limits. Wisdom is held in community, transmitted through relationship, and not safely reducible to documentation. An Eclectic Operations leader draws on all of this. Not as philosophy, but as operational method.

The ABCDE framework of Eclectic Leadership maps directly onto this. A practitioner who can navigate all five dimensions with genuine fluency is demonstrating Eclectic Intelligence in practice. The E cluster alone, covering Ethical, Emotional, Ecological, and Eclectic dimensions, holds more intelligence typologies than most leadership development programmes cover in a year. Eclectic Intelligence is not one item in that list. It is the capacity that allows you to use the whole list well.

A recent Gartner survey of 469 CEOs found that 80% say AI will force operational capability overhauls. (Gartner, April 2026) The overhaul they are describing is not a technology upgrade. It is an intelligence upgrade. And the leaders who will navigate it most effectively are not the ones with the most AI tools. They are the ones who have developed the Eclectic Intelligence to know which tool, human or artificial, the situation actually calls for.

What This Means If You Are Running Operations Right Now

The Troika was built as an Eclectic framework before Eclectic Leadership had a name.

• Your framework needs an outer ring. AI, HI, and Eclectic Intelligence are the operating conditions your Troika is already running inside.

• Your internal stack needs Context below Culture. Read the room before you try to change it.

• Your People arm needs to include Agents. Budget for tokens. Govern accordingly.

• Your Operations team is already your primary culture carrier. Design that role deliberately, or it will happen by accident.

• Your most valuable strategic investment right now is not another AI subscription. It is developing Eclectic Intelligence, in yourself and in your team.

Ready to Implement This in Your Organisation?

If this framework resonates and you want to work through how the Troika of Operational Excellence applies to your specific operational challenges, I work with a small number of organisations each year to do exactly that.

The next cohort is open. We work through the full framework, map it to your current operational reality, and build a practical roadmap for implementation, including how to integrate AI Agents, develop Eclectic Intelligence, and shift from culture-first to context-first operations.

If that sounds like the conversation your organisation needs, the door is open: rononiti.org/solutions

Shehzaad Shams

London, UK

---

Chaos. Confusion. Conflicts. The solution is to be Eclectic.