The work that starts after the demo
The demo always works.The Tuesday after it is the problem.
What needs to change around the technology for it to actually work?
The work that starts after the demo
What needs to change around the technology for it to actually work?
01 / The position
That is the work that starts after the demo. It is unglamorous, it is where adoption fails, and it is almost never anyone's job.
Most conversations about AI resolve upward, into leadership, mindset, and culture. Mine resolves downward, into the operating layer. Who owns this once the pilot ends? Where does the knowledge live? What breaks in the workflow when the tool arrives? Which decisions have to be made before a single person is trained?
I work with learning, HR, and technology teams at that layer, where strategy meets real systems, real constraints, and real people who have to run it on Monday.
02 / The argument
Organizations mistake adoption for capability the way learners mistake familiarity for learning. Completions, attendance, certifications, and login counts all measure contact. Capability shows up later, in the work.
A clean dashboard is not evidence. It shows what was easiest to count, which is a different thing entirely. Someone can finish the course and still need help thirty days later. A team can rate a session highly and keep working exactly as before.
The question underneath every engagement I take is not whether people were trained. It is whether behavior changed, and whether the surrounding system was built to support that change.
AI does not fix ungoverned information. It exposes it.
Then
Learning management
A place to assign and record training.
Next
Learning experience
A place people choose to go and explore.
Now
A learning ecosystem
Learning connected to knowledge, data, and the systems people work in.
The destination
An intelligence ecosystem
Learning, knowledge, work, and technology stop operating as separate systems.
03 / The model
Five phases move ownership to the teams doing the work. A central enablement function carries the early phases, then steps back. Success is measured by what keeps running after it does.
A readiness assessment establishes where capability already sits, so the phases start from evidence, not assumption.
Phase 01
The readiness baseline surfaces the people already doing this work who scored well above their peers. They become the ones a team asks first, so support spreads across many people instead of concentrating in one.
Phase 02
Everyone brings a real business problem and leaves having designed something in response to it. The session builds capability in the room. It is not a demonstration.
Phase 03
Support is in the room but not at the front of it. This is the first honest test of whether capability moved or whether it only looked like it did.
Phase 04
Tiered governance is checked against live use cases, not against a policy document. Rules that survive contact with real work are the only rules worth writing down.
Phase 05
The team operates and improves the work on its own. Review sits with the leaders who understand the context. If it still depends on a central function after this point, the model did not work.
The alternative is a central function that never stops being the bottleneck. That is dependency, not enablement.
04 / What I am writing about
Each one fails because something is missing from the system, not because someone made a bad decision. The book in progress works through all six, and what sits above them.
Learning
"Training creates performance."
The unit of analysis is the system, not the course.
Knowledge
"Information equals knowledge."
Information becomes knowledge only when someone structures, owns, maintains, trusts, and uses it.
Governance
"Technology solves organizational problems."
Technology amplifies the system it enters. The conditions were the decision.
Intelligence
"More AI creates more value."
Value comes from workflow integration and permission, not from tool count.
Performance
"Adoption equals capability."
Adoption is evidence of contact. Capability shows up later, in the work.
Experience
"AI should replace human thinking."
The future skill is not prompt writing. It is discernment.
Above all six sits one more. "AI is a technology initiative."
05 / The series
Working through what implementation requires, from inside real organizations instead of from outside them.
Every piece answers some version of the same question: what does this actually take, and who has to own it? Names of tools, systems, and organizations are changed. The problems are not.
06 / Background
This rests on doctoral research into workplace learning ecosystems and on years of enterprise-scale implementation inside a large matrixed organization. I built the systems, implemented them, trained the people who took them over, ran the follow-up teaching, and wrote the governance.
I did not observe this work from the balcony. I did it from the floor.
Years in Learning and Development
Doctor of Instructional Design and Technology
Pillars the work is built on
Where this continues
I write about it while it is still in progress. LinkedIn is where that happens.