Making change stick
These insights come from my work inside organizations navigating change, from AI adoption to broader shifts in how teams operate and build capability. Through conversations with employees, teams, and leaders, I’ve seen consistent patterns in what enables progress and what slows it down.
The articles and frameworks on this page capture those patterns and offer practical ways to assess, measure, and sustain change over time.
THE REALITY CHECK
What AI adoption actually looks like inside organizations
AI adoption is not a curve, it is a patchwork
AI adoption rarely moves evenly. Confident users often sit alongside unchanged workflows and cautious peers, creating the appearance of progress without a shared organizational baseline.
ADOPTION | DISTRIBUTION
Technical fluency as the strongest predictor of AI adoption
Adoption tends to follow comfort with digital tools and ambiguity. Age, title, and tenure matter far less than hands-on fluency and willingness to experiment.
CAPABILITY | SIGNAL
Training is not the bottleneck, workflows are
Learning about AI is rarely enough on its own. Adoption becomes reliable when AI is woven into how work already gets done.
WORKFLOWS | ENABLEMENT
RISK AND TRUST
Why caution is rational and what actually enables confidence
AI is mostly used as a bright intern
Much of today’s value comes from assistive use. AI helps with structure and momentum, while humans keep judgment and ownership.
COLLABORATION | BEHAVIOR
High-stakes work changes everything
Where accuracy and accountability are critical, AI use evolves more carefully. Slower adoption often reflects thoughtful risk management rather than hesitation.
ACCURACY | TRUST
People are not resisting AI, they are protecting themselves
Caution is usually about responsibility, not attitude. When expectations and safeguards are clear, people are more willing to engage.
SAFETY | BEHAVIOR
FROM USE TO SCALE
How organizations move from experiments to systems
Guardrails enable adoption, not fear
Clear boundaries make experimentation safer. When people know what is allowed and what requires review, AI use becomes more consistent and sustainable.
BOUNDARIES | SCALE
Large files are the silent adoption killer
Large documents, complex spreadsheets, and fragmented inputs quietly cap adoption. When AI cannot reliably ingest the materials people actually work with, use remains shallow regardless of intent or skill.
OPERATIONS | READINESS
Agents come last
Automation works best after foundations are in place. When workflows and trust are established first, agents become a natural next step rather than a risk.
