Point of View
Convictions — and why they matter.
These are not negotiable positions. They are convictions built through years of building, deploying, and operating AI systems inside complex organizations.
Memory Before Orchestration
Here is the uncomfortable truth about enterprise AI: most of it has amnesia.
An organization deploys a chatbot. It answers questions. But tomorrow, it has forgotten everything it learned from today's conversations. An automation pipeline processes a thousand documents. Next week, it cannot recall a single pattern from that work.
This is not a feature gap — it is a structural failure. And it is the reason most enterprise AI initiatives plateau after the initial excitement fades.
Fortune 100 companies invest millions in AI tooling while completely ignoring the memory layer. They buy the engine but forget to build the road. Every interaction starts from zero. Every insight evaporates. Every lesson is lost.
Memory is not a nice-to-have. Memory is the foundation. Without persistent organizational memory, AI cannot learn from its own operations. It cannot compound its understanding. It cannot evolve from tool to trusted system.
Memory before orchestration. Always.
IT Is Not the Right Steward of AI Transformation
This will make some people uncomfortable. Good.
IT departments are built to maintain systems, manage infrastructure, and reduce risk. These are essential functions. But they are the wrong orientation for AI transformation.
AI transformation is a business strategy problem, not an infrastructure problem. It requires understanding customer behavior, revenue mechanics, operational workflows, and competitive dynamics. It requires someone who can sit in a boardroom and connect a machine learning model to a P&L statement.
There are rooms where IT leaders present AI roadmaps that are technically sound and strategically meaningless. Beautiful architectures that solve no business problem. Perfectly governed data that generates no insight.
The best AI transformations are led by operators — people who understand both the technology and the business context deeply enough to build systems that create measurable value. Not IT administrators. Not consultants who have never shipped production code. Operators.
This is not an insult to IT professionals. It is a recognition that AI transformation demands a different kind of leadership — one that speaks business outcomes fluently and builds technology that delivers them.
Practitioners Trump Theorists
The enterprise AI advisory space is full of people who have never built anything.
They have read the papers. They have attended the conferences. They can draw impressive architecture diagrams on whiteboards. But they have never sat in front of a production system at 2 AM trying to figure out why the orchestration pipeline is dropping 30% of its context.
There is a kind of knowledge that only comes from building — from deploying AI applications inside real organizations with real constraints, real politics, and real consequences. This knowledge cannot be summarized in a framework deck. It lives in the scar tissue of failed deployments and the muscle memory of successful ones.
Advising executives on AI strategy should draw from applications built, deployed, and maintained inside complex organizations — not from case studies. Knowing what breaks, what scales, and what the theory leaves out is the difference.
If your AI advisor has never written a deployment script, find a different advisor.
Autonomous Orchestration Is Inevitable
The question is not whether organizations will deploy autonomous AI systems. The question is whether those systems will be trustworthy.
Trusted Autonomy is the end state of a maturity model refined through years of practice. Start by delivering. Measure what gets delivered. Learn from those measurements. Remember what is learned. Only then expand.
Most organizations want to skip to autonomy. They see the demos. They read about agents that can book meetings and write code and manage workflows. And they think: we need that.
But autonomy without trust is chaos. An autonomous system without memory is a liability. An autonomous system without governance is a lawsuit.
The organizations that will win the AI era are not the ones that deploy autonomous agents fastest. They are the ones that build the infrastructure — the memory, the measurement, the governance — that makes autonomy safe, auditable, and genuinely valuable.
Not the flashy part. The foundational part. The part that determines whether an AI investment compounds or collapses.
The Enterprise AI Conversation Is Broken
Listen to what the enterprise AI conversation sounds like right now: "We need a GenAI strategy." "Let's build a chatbot." "Can we use AI to automate our reports?"
These are not strategies. These are shopping lists.
A real AI strategy starts with a different set of questions: What does your organization know? Where does that knowledge live? How does it flow between people, systems, and decisions? What would it mean for your organization to actually remember what it learns?
Most enterprises cannot answer these questions. And that is exactly the problem.
The conversation needs to shift from "What AI tools should we buy?" to "What intelligence infrastructure do we need to build?" From individual applications to organizational capability. From point solutions to compounding systems.
The conviction here is straightforward: the enterprises that treat AI as an infrastructure investment — not a tool purchase — will create a widening competitive advantage. The rest will spend the next decade wondering why their AI investments never delivered on their promise.
Not a popular position with vendors. Not easy to execute. But correct.
If this resonates, we should talk.
The AI Readiness Briefing is a personalized assessment of where your organization stands — and what it takes to build real intelligence infrastructure.
Schedule Your Briefing