Enterprise leaders spend heavily on AI. Pilots launch. Training portals go live. And yet, most organizations struggle to convert AI Implementations into sustained business impact.
In this episode of ETMA Tech Talk, Melissa Reeve, author of Hyperadaptive: Rewiring the Organization to Become an AI-Native Enterprise, explains why. Drawing on enterprise research and real-world implementation work, she makes a clear case that AI failure rarely comes down to the model or the tool. It comes down to how the organization is wired.
According to research from the RAND Corporation, 80 percent of AI initiatives fail to meet expectations. Not because leaders lack ambition, but because the systems that govern decisions, learning, funding, and execution were never designed for an AI environment that changes every three to six months.
This conversation reframes the problem. Instead of “Which AI should we buy?” This discussion focuses on something more fundamental. How does an enterprise rewire itself to keep learning while technology keeps evolving and changing every day?
AI Is Not Static Software. Treating It Like One Guarantees Friction.
One of the most practical insights in this episode is also one of the most overlooked. AI is not a tool you implement and learn once. It is an exponentially evolving capability.
Models, interfaces, guardrails, and failure modes shift constantly. Most training delivered today is obsolete within months. Expecting every employee to independently track that pace is unrealistic, especially when they already have full-time jobs.
Reeve argues that enterprises need to stop thinking about AI training as a “one and done” event and start treating it as ongoing. A small subset of the organization must be responsible for keeping up with current distributing updates that matter in short, usable bursts. Not once a year.
Without that structure, organizations see the same pattern repeat itself. A small group of power users pushes ahead. A large middle group uses AI lightly for email and summaries. A meaningful portion opts out entirely. Productivity fragments instead of compounding.
The Hidden Cost No One Budgets For: Human Infrastructure
Most AI budgets focus on tools. Licenses, cloud usage, and vendors. What gets underfunded is the infrastructure that supports the humans expected to use those tools responsibly.
Reeve draws a parallel to the rise of the PC. When computers entered the enterprise, companies did not simply hand them out and hope for the best. Help desks, IT support, security practices, and usage standards were rolled out alongside the technology.
AI demands the same treatment. Leaders need clear answers to practical questions that surface every day. Who do you call when the model hallucinates? Who decides whether sensitive data can be uploaded? Who validates that an AI output can be trusted in a regulated workflow?
When those questions go unanswered, risk avoidance sets in. Adoption stalls quietly. On paper, the AI rollout exists. In practice, it never embeds into how work actually gets done.
Linear Enterprises vs AI-Native Competitors
A core theme of Hyperadaptive is the structural disadvantage large enterprises face when competing with AI-native organizations.
Traditional enterprises operate linearly. Strategy flows downward. Execution moves across layers. Work travels through handoffs. Decisions take time.
AI-native organizations compress both hierarchy and delivery. Decisions move faster. Skills flatten. Fewer handoffs are required to move from idea to outcome.
The risk for large enterprises is not that they fail to buy AI. It is that they fail to close this structural gap. Reeve’s work focuses on helping linear organizations move incrementally toward an AI-native operating model, without blowing up stability or governance along the way.
That transition requires intentional support structures, not heroic individual effort.
Why AI Use Cases Stall Before They Scale: The FOCUS Framework
One of the most actionable parts of the discussion is the FOCUS framework Reeve uses to prioritize AI initiatives before they consume budget and credibility.
FOCUS stands for:
Fit
Does the use case align directly with organizational strategy?
Organizational pull
Will people actually use it, or will it sit unused once the novelty fades?
Capabilities
Does the organization have the skills to build and support it today?
Underlying data
Is the data foundation strong enough for the use case to work reliably?
Success metrics
Can value be demonstrated in terms the business recognizes?
In a world where AI can be applied almost anywhere, prioritization becomes the discipline that separates momentum from noise. The framework helps leaders avoid random acts of AI and focus investment on areas where outcomes can be proven.
Governance That Moves at the Speed of AI
Governance is where many AI initiatives quietly fail. Not because governance exists, but because it is designed for a slower world.
Standing councils that meet quarterly cannot keep pace with models that evolve monthly. Static policy documents buried on an intranet do not help employees make real-time decisions.
Reeve introduces the idea of dynamic governance. Governance that is embedded inside the AI tools employees already use. Policies that update continuously. Guidance delivered in context, at the moment of decision.
This approach also creates a feedback loop. Leaders can see what employees are asking, where confusion exists, and which use cases are emerging faster than policy anticipated. Governance shifts from gatekeeping to enablement, without abandoning control.
ROI, FinOps Pressure, and the J-Curve Reality
CFOs and FinOps leaders face immediate pressure to justify AI spend. That pressure is valid. It also creates tension.
Reeve describes AI ROI as a J-curve. There is often a short-term dip as people learn, workflows change, and processes are rewired. Early outputs may look impressive, but output volume is not the same as business impact.
The shift that matters is from measuring activity to measuring outcomes. Faster decision cycles. Reduced handoffs. Shorter time from insight to action. These benefits compound over time, but only if leadership allows room for learning before demanding optimization.
What Real Success Looks Like: The Moderna Example
Among the examples discussed, Moderna stands out. Their stated goal was ambitious. Deliver 15 new drugs in five years using AI.
What mattered was not just the goal, but how the organization mobilized around it. Moderna identified AI leads through internal prompting competitions, supported them systematically, and created active communities of practice with thousands of weekly participants.
Middle managers played a critical role. Research cited during the conversation highlights that transformation succeeds most often from the middle out, where strategy meets execution. These leaders translate vision into operational reality and create capacity for learning on the ground.
Contrast that with top-down mandates driven by fear. Cultural resistance rises. Adoption drops. The technology becomes symbolic rather than operational.
Learning Loops as a Competitive Advantage
AI rewards organizations that learn faster than their competitors. Reeve describes an AI learning flywheel built around four stages: spark, spread, scale, and sustain.
Initial curiosity matters. So does local relevance. Learning scales when trusted peers teach applied use cases inside real workflows. It sustains when dedicated activation hubs monitor change, update guidance, and distribute learning continuously.
This system allows large enterprises to compete effectively with smaller, faster firms by turning scale into an advantage rather than a drag.
The Corporate Habit That Holds AI Back the Most
When asked which outdated practice most undermines hyperadaptivity, Reeve points to the annual budgeting cycle.
Static funding encourages territorial behavior, discourages experimentation, and slows response to change. AI thrives in environments where funding can adjust dynamically, supporting value streams and innovation loops alongside the stable core of the business.
Money drives behavior. Changing how it flows changes how organizations move.
The Skills Leaders Need More Than Ever
Two traits surfaced repeatedly in the conversation. Curiosity and critical thinking.
Curiosity fuels exploration and keeps leaders engaged as tools evolve. Critical thinking ensures AI outputs are challenged, contextualized, and improved rather than blindly accepted.
AI elevates capability. It does not replace judgment.
A Final Thought for Enterprise Leaders
If AI offers unprecedented power, it also demands responsibility. The same systems that accelerate value creation can amplify risk, inequity, and environmental impact if left unchecked.
Hyperadaptive organizations do not chase every new capability. They build systems that help people learn, decide, and act well as change accelerates.
That is the real transformation.
Learn More
Explore Melissa Reeve’s work, assessments, and enterprise research, including “Where Do You Stand with Your AI Foundation?” diagnostic.
Listen to the full ETMA Tech Talk conversation for deeper context and implementation insight.
- Connect with Melissa Reeve https://www.linkedin.com/in/melissamreeve
- Connect with https://hyperadaptive.solutions
- Where Do You Stand with Your AI Foundation https://hyperadaptive.solutions/discover
