The Great AI Divide — Part II
Your AI Initiative is Failing Because You Forgot How to Build Things
I'm a CTO and founder with nearly two decades of experience driving growth and transformation through technology. At Stronghold Investment Management, I led the development of a systematic real asset trading platform and modernized everything from Salesforce strategy to custom cloud-native infrastructure. My background spans commercial real estate, e-commerce, and private markets — always focused on delivering innovation, velocity, and meaningful business outcomes. I hold a PhD in Theoretical & Computational Biophysics and was recognized as a Google Developer Expert in Cloud. I build high-trust, high-output teams. I’ve rebuilt broken cultures, hired top-tier engineers, and helped early-stage and PE-backed companies scale with confidence. System modernization is my specialty — not just upgrading software, but aligning teams and infrastructure with what the business actually needs. Currently, I lead client engagements through Heavy Chain Engineering and am building Newroots.ai, an AI-driven relocation advisory platform.
In the last post, I talked about the “Great AI Divide” separating the 5% of companies getting massive value from AI from the 95% who are stuck with a 95% pilot failure rate. The MIT report that uncovered this depressing statistic also, unintentionally, revealed the cause. And it’s something I’ve been shouting from the rooftops.
Your AI initiative is failing because you’ve forgotten the fundamentals of running a business. It’s that simple.
I’m not just talking about software. I’m talking about the disciplined process of building anything. And a huge part of that discipline is something every MBA has studied: change management. We have decades of knowledge about how to introduce new things into an organization, yet with AI, it’s like we’ve thrown it all out the window. The MIT report notes that most enterprise AI tools fail because they are “brittle,” “overengineered,” or “misaligned with actual workflows.”
Of course they are. Because the teams building them are treating AI like a magical incantation. They throw a vague request at the machine and are shocked when the output is garbage. As I wrote recently, we’re accidentally reinventing software engineering because AI is forcing us to be rigorous again. It’s forcing us to remember that planning matters.
The Real Moat: Your Process
The Wharton report shows that the most successful adopters of AI are in IT, Purchasing, and Procurement. Marketing and Operations, surprisingly, lag behind. But here's what's interesting: it's not about the people or the function. It's about the maturity of the processes they inherit.
IT and Procurement operate in environments where decades of compliance, audit trails, and vendor management have forced rigorous documentation. You can't procure a million dollars of hardware without a detailed plan. You can't deploy a new software system without clear specifications. These disciplines didn't choose to be more structured—they were forced to be by regulation, risk management, and the sheer cost of failure.
Marketing and Operations, by contrast, often operate in environments that reward speed and creativity over documentation. That's not a weakness—it's a feature. But it means they don't have the same forcing function to create the machine-readable processes that AI needs to thrive.
This is the secret the 5% know. The value isn't in the AI model; it's in the quality of the instructions you give it. As I've said before, businesses with strong business processes will see stronger acceleration than those that don't, because they can translate that process into machine-readable specs for AI to implement.
The MIT report found that internal AI builds fail twice as often as projects with external partners. This isn't because partners have better AI. It's because working with a partner forces you to do the one thing you've been avoiding: think clearly about what you actually want. You have to write a statement of work. You have to define deliverables. You have to create a plan.
When you're just playing with a tool internally, it's easy to be lazy. You can get away with vague goals and sloppy thinking. The result? Science projects and wrappers that go nowhere.
Stop Trying to Make AI Smarter
The most successful new AI tools—the ones that are actually getting traction with developers—aren’t about better models. They’re about better workflows. Tools like Task Master, AgentOS, and GitHub’s SpecKit are all designed to solve a human problem, not a technical one. They provide frameworks for clear thinking, proper planning, and detailed specifications.
They are, in essence, guardrails against our own worst instincts. They force us to move from “throw directions at AI and hope” to “create clear specifications that AI can reliably implement.”
This is the shift I see in my own work. When I train a team, we don’t play with prompts. We build frameworks. For a marketing team, we create a brand voice document. For an engineering team, we write executable specs. We give the AI context, structure, and a clear goal. And suddenly, it stops being a disappointing toy and starts being an incredible accelerant.
AI doesn’t magically fix a broken process; it just exposes it faster.
This brings me back to my initial point: something like 80% of all standard initiatives fail anyway. That’s not a coincidence. It’s the root cause. AI doesn’t magically fix a broken process; it just exposes it faster. The 95% failure rate for AI pilots isn’t an AI problem. It’s a symptom of a deeper organizational sickness: the belief that this new ‘magic wand’ exempts you from the hard work of leadership, planning, and disciplined change management. It’s the ultimate abdication of thinking, and it’s happening at scale.
If you want to cross the GenAI Divide, stop looking for a better AI. Look for a better process. Stop trying to make the AI smarter. Start making your specifications clearer. The future of this technology isn’t about replacing human thinking. It’s about elevating it. It’s about finally giving us the leverage to put our grey matter on what actually matters.
In the final post, we'll discuss how to cultivate the right talent and mindset to thrive in this new era of accountable acceleration.
