Last week, I wrote that AI is breaking the org chart permanently — that 40% of middle management roles are disappearing right now because organizations can’t afford the latency hierarchies create at agent speed.
The responses fell into two camps. Half said, “We see this coming. We’re hiring AI leaders and forming transformation committees.” The other half asked, “Why can’t we change even when we know we need to?”
Both groups overlook the underlying problem. You can hire every AI expert on LinkedIn. You can form cross-functional tiger teams. You can mandate agile practices and preach cultural transformation. And you’ll still fail — not because your people resist change, but because your organization’s processes and values won’t let you redraw the boundaries needed for new work patterns.
Clayton Christensen explained this 20 years ago. His Resources-Processes-Values (RPV) framework shows why organizations can’t evolve even when they anticipate disruption. The telegraph org chart was a Resources problem. The inability to redesign it is a Processes and Values problem. And AI isn’t breaking your org chart — it’s exposing that your organizational DNA won’t permit the surgery required.
The RPV Diagnostic: Why Hiring Doesn’t Equal Capability
Christensen’s framework explains how organizational capabilities migrate through three stages: Resources → Processes → Values.
Resources are the easiest to change. People, technology, cash, and partnerships —these can be acquired. When a startup needs AI capability, it hires data scientists. Problem solved.
Processes are harder. These are the patterns of interaction, coordination, communication, and decision-making that transform inputs into outputs. Processes embed themselves deep in organizational muscle memory — routing approval through five layers, requiring cross-functional sign-off, and holding weekly status meetings. Processes are how work gets done, regardless of what the strategic plan says.
Values are hardest. These are the criteria by which decisions are prioritized — what gets resourced, what gets delayed, what gets killed. Values determine whether you prioritize predictability over speed, consensus over autonomy, or risk mitigation over learning velocity.
Here’s the problem: capabilities migrate from Resources to Processes to Values over time. What starts as “hire smart people” becomes “follow these coordination patterns,” which hardens into “we don’t work that way here.”
You can hire AI leaders. But if your processes route decisions through five approval layers and your values prioritize predictability over adaptive learning, nothing changes. The new resources get absorbed into existing processes, which are protected by deeply embedded values.
That’s why 46% of AI pilots fail. Organizations treat AI adoption as a Resources problem — hire talent, buy tools, fund projects. But the failure happens at the Processes and Values level. Your organization’s DNA rejects the transplant.
Organizational Boundaries Protect Processes and Kill Innovation
Why do organizational boundaries persist even when they’re obviously misaligned with current work patterns?
Because boundaries exist to make processes efficient. When communication traveled by telegraph, organizing by geography made perfect sense. Regional managers coordinated work within their physical boundaries because crossing boundaries meant days-long delays. The structure optimized for the constraint.
Now communication is instant. But the boundaries remain — not because anyone thinks they’re optimal, but because existing processes depend on them. Performance reviews, budget allocation, resource planning, career progression — hundreds of embedded processes assume the current structure. Changing boundaries means redesigning all those processes simultaneously.
And that triggers the Values layer. Organizations that prioritize stability over adaptability, efficiency over effectiveness, risk mitigation over learning — their values protect existing boundaries because those boundaries make current processes run smoothly.
This is the core insight from the Healthcare.gov disaster. The project didn’t fail due to a lack of resources. It failed because organizational boundaries impeded the interaction patterns required for complex-system integration. Teams were separated by departmental walls. Processes required sequential handoffs. Values are prioritized over outcomes in the procedure.
No additional resources could resolve a Processes and Values problem.
The Heavyweight Team Solution: When to Redraw Boundaries
Andy Boyd’s frameworks provide the diagnostic tool. When new work requires different interaction patterns than existing processes support, you need heavyweight teams.
Lightweight teams work within existing boundaries. They borrow resources from functional departments. Coordination happens through normal processes. This works when the work fits established patterns.
Heavyweight teams redraw boundaries. They extract resources from functional homes and co-locate them under dedicated leadership, with autonomous decision-making authority. This works when the work requires sustained interaction patterns that existing processes prevent.
The diagnosis is straightforward: Does success require frequent coordination between functions? Do decisions need to happen faster than current approval processes allow? Is learning velocity more critical than efficiency?
If yes, existing boundaries are blocking required interaction patterns. Redraw them.
That’s what’s happening with the 40% management reduction. Organizations adopting AI at scale are discovering that when 2–5 humans supervise 50–100 agents, old boundaries physically prevent required interaction patterns.
Heavyweight team logic, applied at an organizational scale.
Why Change Management Fails: Treating Process Problems Like Communication Problems
Most change management programs fail because they misdiagnose the problem. They assume resistance comes from poor communication, insufficient training, or cultural inertia.
But resistance is often rational. When people say “this won’t work here,” they’re often correct — not because the idea is flawed, but because existing processes and values make it impossible to execute.
The secret to successful AI-driven process redesign isn’t better communication. It’s an explicit process redesign. You can’t overlay new work patterns onto old processes and expect success. You must redesign the processes themselves—and be prepared for the value conflicts that result.
This is where most operating model redesigns fail. Organizations focus on drawing new org charts without redesigning the processes that make work happen or addressing the values that determine what gets prioritized.
New boxes, same processes, identical values. Nothing changes.
The race to redesign isn’t about organizational aesthetics. It’s about whether your processes and values permit the interaction patterns AI requires.
The AI Forcing Function: A 2×2 Diagnostic
AI creates a forcing function because the speed mismatch isn’t incremental — it’s exponential. You can’t optimize telegraph-era processes to operate at millisecond speed. The gap is measured in orders of magnitude.
Barry O’Reilly’s framework illustrates the strategic choice: organizations either redesign processes and values to match AI execution speed, or they become non-competitive relative to organizations that do.
Here’s a simple 2×2 diagnostic for when to redraw boundaries:
Axis 1: Interaction Frequency
- Low: Work happens within existing functional boundaries
- High: Success requires sustained cross-functional coordination
Axis 2: Decision Speed
- Slow: Existing approval processes match work tempo
- Fast: Decisions need to happen faster than current governance allows
Low interaction + Slow decisions: Keep existing boundaries. Optimize current processes.
Low interaction + Fast decisions: Distribute authority within existing boundaries. Lightweight teams with clear decision rights.
High interaction + Slow decisions: Co-locate teams but maintain existing governance. Cross-functional squads reporting through normal channels.
High interaction + Fast decisions: Heavyweight teams. Redraw boundaries, redistribute authority, redesign processes. This is the agentic organization zone.
When 2–5 humans supervise 50–100 AI agents making thousands of decisions daily across multiple value streams, you’re in the top-right quadrant. Old boundaries don’t just slow things down — they make execution physically impossible.
That’s the forcing function. Not ideology. Mathematics.
The Practical Reality: Process Redesign, Not Persuasion
If your organization can’t evolve despite seeing the problem coming, stop treating it as a communication challenge. Start treating it as a Processes and Values problem.
Three diagnostic questions:
What processes depend on current organizational boundaries? Map them explicitly. Performance management, budget allocation, resource planning, career progression, approval workflows — which processes assume the current structure? Changing boundaries means redesigning all of them simultaneously. Are you prepared for that scope?
What values protect those processes? Does your organization prioritize efficiency over effectiveness? Predictability over adaptive learning? Risk mitigation over velocity? Consensus over autonomy? Those values will defend existing processes even when those processes block required capabilities.
Where do new work patterns require different boundaries? Use the 2×2 diagnostic. Where does success require high interaction frequency plus fast decision speed? Those are your heavyweight team candidates. Redraw boundaries there first. Prove the model works. Expand systematically.
The telegraph org chart diagnosis was accurate. The AI breaks the org chart article, documenting the collapse. This article explains why organizations can’t evolve even when they see it coming: because organizational capabilities migrate from Resources to Processes to Values, and disruption occurs faster than that migration can be reversed.
You can hire AI leaders. But if your processes route through five approval layers and your values prioritize stability over adaptation, those leaders will execute within existing constraints — or leave.
The fix isn’t better communication. It’s an explicit process redesign, supported by values evolution and enabled by boundary redefinition.
AI didn’t break your processes. It exposed that your organization’s DNA won’t permit the changes required to survive.
Evolution isn’t optional. The forcing function is mathematical. The gap between diagnosis and surgery is measured in Processes and Values, not in Resources.
Ralph Jocham is Europe’s first Professional Scrum Trainer, co-author of “Professional Product Owner,” and contributor to the Scrum Guide Expansion Pack. As an ICF ACC certified coach, he works with organizations to build Product Operating Models where strategic clarity, operational excellence, and adaptive learning create measurable competitive advantage. Learn more at effective agile.