For the first decade of a career that spans product management and data analytics, doing the work meant diving directly into the mechanics of corporate data. Writing database queries that pulled raw figures out of sprawling corporate systems, building statistical models to predict what customers would inevitably want to buy, and shipping data pipelines that moved critical information smoothly between underlying business systems consumed the daily routine. Over the years, those foundations allowed for the building and scaling of analytics teams at major national brands like Best Buy and Target, closely studying how modern consumers shop and what local stores should stock to meet shifting demands.
Today, that trajectory continues at Lowe’s, the Fortune 100 home improvement retailer, where the focus centers on leading enterprise artificial intelligence transformation. Yet the ultimate objective across the retail landscape has never been simply to sell artificial intelligence or push flashy new technology for its own sake. Instead, the real goal is to harness these advanced computational tools to deliver genuinely useful expertise precisely at the moment a customer needs it most. In retail and other customer-facing industries, virtual assistants can help people address everyday, high-friction questions—such as figuring out how to repair a persistent leaky faucet—while thoughtfully guiding them toward the most relevant products, services, or logical next steps.
As these capabilities become increasingly common fixtures of the digital economy, traditional technology roles are undergoing a profound transformation. The work is no longer limited to simply building AI systems from the ground up; it now firmly includes defining how those systems operate in the wild. Professionals must determine which decisions artificial intelligence can make autonomously, when systems must immediately escalate a complex scenario to a human expert, and which operational actions must remain strictly off-limits.
That fundamental shift—moving away from merely building AI systems toward actively governing them—is coming for anyone who is held accountable for what such machines ultimately produce. This evolution does not apply to the casual user casually typing prompts into a consumer chatbot. Rather, it targets the engineers, product managers, data analysts, and business operators who ultimately sign off on the work a machine has drafted before it hits the production environment.
This evolving landscape forms the central subject of The Enterprise Brain, a book coauthored to examine the modern corporate reality. Within its pages, this transition is defined as the "governor shift"—a transition away from executing tasks directly yourself and toward setting the intent, core principles, and rigid boundaries within systems that execute those tasks on your behalf. Modern business operators might not write the underlying code, but they are the ones who must decide which pricing exceptions an AI agent is permitted to approve independently and which scenarios require mandatory human escalation. That is what governing truly means in practice.
Despite the sweeping enthusiasm surrounding emerging technologies, a sobering 2025 report from the MIT Media Lab’s Project NANDA revealed a striking disconnect in the market. The research found that despite an estimated US $30 billion to $40 billion pouring into enterprise generative-AI investments, the vast majority of organizations included in its dataset had not yet demonstrated a measurable profit-and-loss impact. The report estimated that only about 5 percent of integrated pilots were actually generating substantial, verifiable business value. This stark reality underscores just how difficult it remains for modern corporations to successfully move from initial experimentation to scaled, long-term business outcomes.
Researchers characterized this widespread industry pattern as the GenAI Divide, a framing adopted in recent literature to describe the gap between aspiration and realization. Companies caught on the wrong side of the divide rarely lack the necessary technology; in fact, they typically utilize the exact same foundational models deployed by the 5 percent of organizations achieving winner-circle status. What they lack instead are the people capable of directing those complex systems and standing firmly behind the resulting outputs.
The Identity Question and the Scarcity of Judgment
When discussing this operational shift with professionals across industries, the primary source of resistance is rarely technical in nature. More often than not, the hesitation stems from a deeper crisis of identity: If the artificial intelligence is doing the actual doing, what is left for the human professional to do?
Observing capable, highly trained people freeze up when confronted with that exact question is common, and it is a challenge that reflective leaders have had to pose to themselves as well. Upon closer inspection, however, direct execution was never truly the core of the job. Execution was merely the vehicle through which professional judgment was expressed.
Far from diminishing the value of human input, artificial intelligence has actually elevated judgment into the scarcest resource within the modern organization. For the first time in corporate history, one person’s well-articulated judgment, codified and structured correctly, can guide thousands of individual business decisions each and every day.
Moreover, judgment has an overlooked twin that professionals discuss far less frequently: taste. Judgment tells you whether a given answer is fundamentally sound and compliant with corporate policy. Taste, on the other hand, tells you whether the initial question was actually worth asking in the first place, and which of a hundred equally defensible options should be offered to the customer. A machine will happily and instantly generate all one hundred options, but it cannot tell you which one is truly the best fit for the human context.
Navigating this transition successfully rewards the exact instincts that many technical professionals and engineers already possess: rigorous systems thinking, extreme precision regarding requirements, and radical honesty about potential failure modes. The tools have changed dramatically, but the foundational discipline has not.
Governing is the exact domain where taste and judgment cease to be mere soft corporate buzzwords and transform into the actual work itself. The professionals who treat the discipline through that lens, rather than viewing it as a step away from traditional engineering, will ultimately define the parameters of their profession in the age of artificial intelligence.
