Across industries, a quiet but costly problem is compounding. Capital is being poured into “transformation” that does not transform. Operating model redesigns stall before they scale. Capability programs light up dashboards, then fade. The result is a systemic misallocation of resources, a drag on competitiveness that boards can no longer ignore.
The reason is not a lack of effort. It is a failure of focus. Most leaders are managing the wrong problem. They are caught in the Surface Wave.
The Surface Wave: An Illusion of Progress
The Surface Wave is the visible, tactical, immediate activity around artificial intelligence (AI). It is the parade of pilots, the vendor roadshows, the urgent mandates for AI literacy, the proliferation of tools with enthusiastic early usage. It feels like progress. It looks like progress. On a project dashboard, it is a field of green. Yet the fundamentals of the business remain the same. Decision cycles do not shorten. Cost structures do not bend. Authority flows do not change. What you get is motion without progress. Success stories accumulate while organizational fatigue rises. Launches are rewarded, announcements are celebrated, and the scoreboard does not move.
The Undercurrent: The Real Engine of Transformation
The real story, the one that will decide winners and losers over the next decade, sits beneath the surface. I call it the Undercurrent. It is the strategic redesign of how the organization actually functions. It is not about tools. It is about power, decisions, and control.
You can see the Undercurrent in three patterns that are already emerging:
- The de-layering of management. As AI systems take on monitoring, escalation, and routine decisions, middle layers of approval lose purpose. Authority drifts toward the edge, where the work lives. Organizations that cling to legacy hierarchies slow themselves into irrelevance.
- Distributed capability ownership. A head of Sales, equipped with AI for analytics, pipeline quality, enablement, and compensation modeling, can own the performance lifecycle end to end. Centralized service teams no longer hold a monopoly on improvement.
- AI gains a voice in the room. In forward-leaning firms, systems do not just feed analysts. They surface intelligence directly into leadership forums, and in doing so, they bypass traditional chains of command. Leaders who treat that as gadgetry miss the point. It rewires how strategy is made.
The Disconnect: Why Leaders Get Trapped
The Surface Wave consumes attention because it is noisy and easy to count. The Undercurrent is harder to see and harder to fake. It demands choices that change who can act, what evidence is required, and how often judgment is exercised. It asks leaders to give up some familiar control in order to gain speed, clarity, and resilience.
This is where the disconnect appears. AI is entering the enterprise from the bottom up. Employees are already using tools. Leaders focus on tool adoption, while the organization is being reshaped by a force they have not named. The result is a set of failure patterns that repeat across sectors.
Pilot theater is the first. Demos and proofs of concept accumulate because they are safe to sponsor and easy to applaud. They rarely graduate to production because no one has decided who will make decisions differently once the pilot works. Prompt school without redesign is the second. Thousands of people learn to use new tools, then return to old workflows. Skill exists, the system does not. Data fishing is the third. Sophisticated models and dashboards are built on weak data foundations. Without clear ownership, traceability, and trust, the outputs are ignored when stakes are high. None of these are technology problems. They are leadership problems.
The New Mandate: From Tool Deployer to Organizational Architect
The mandate has shifted. Your job is no longer to deploy tools. Your job is to re-architect the enterprise for a world in which intelligence and authority are distributed differently.
Start with decision rights. Make them explicit: Who, and what, has the authority to act in each domain. What minimum evidence is required. When an AI system can proceed without approval, and when it must escalate. If this is not clear, the organization will oscillate between paralysis and misuse.
Next, build governance that can scale. Governance is not bureaucracy. It is the safety system that makes innovation repeatable. Boards should expect runtime oversight, clear incident response, and change control that can handle autonomy without chaos.
Finally, cultivate the courage to act. The most powerful systems will surface uncomfortable truths: pet projects that underperform; strategies that misallocate; processes that look efficient until the full cost of rework is measured. The limiting factor is rarely the model. It is the willingness to act on what the model reveals.
What the Undercurrent Looks Like in Practice
These choices are not abstract. Picture an AI-native strategic intelligence unit. In the Surface Wave, analysts get better tools for research and synthesis. Useful but incremental. In the Undercurrent, the system becomes a participant. Insights flow directly to each decision-maker, not just through a manager. Meeting agendas are shaped by what the system has found, not by the loudest voice in the room. Accountability follows the information, not the hierarchy. That is not a software upgrade. That is a redesign of how the organization thinks and decides.
Or picture a commercial leader with full lifecycle ownership. In the Surface Wave, the sales team adopts a new assistant for e-mails and summaries. In the Undercurrent, the business owns hiring profiles, onboarding playbooks, enablement cadences, and compensation levers, all informed by shared data and tested in short cycles. Finance shifts from quarterly post-mortems to continuous calibration. Enablement stops being a service desk and becomes an internal market for capability building. That is what distributed ownership looks like when it moves the scoreboard.
A New Posture for Leadership
Leaders often ask for a checklist at this point. Checklists are part of the Surface Wave. The Undercurrent requires a different posture. Set a simple rule for yourself and your board. Reward redesign, not announcements. Fund changes that move decision rights, not projects that add tools. Ask for evidence that cost curves are bending, decisions are faster, and accountability is closer to the customer. Measure whether people trust and adopt the new ways of working. An AI initiative that cuts costs but erodes trust is a net loss. One that improves both efficiency and confidence is a real gain.
People readiness is as critical as system readiness. Employees need to feel supported, not surveilled. Managers need to learn how to oversee autonomy without fearing it. The workforce needs practical AI literacy so teams can question, calibrate, and work alongside systems responsibly. Without adoption and trust, even well-governed designs fail to take root.
A final word on pace. Breakthroughs may accelerate timelines, but betting your strategy on speed is risky. The safer move is preparation, not prediction. Overreach destroys credibility. When leaders echo vendor timelines too confidently, they stake their reputation on promises they do not control. Frame autonomy as a possibility to prepare for, not an inevitability to promise.
Transformation will remain hard. It does not have to remain wasteful. The winners of the next decade will not be the organizations that surf the Surface Wave the longest. They will be the ones that master the Undercurrent first, the ones that redesign decision rights, governance, and accountability before the tide turns. The rest will exhaust themselves in pilots, workshops, and dashboards, and will discover too late that their AI strategy was an illusion of progress.
