Seats get purchased, a lunch-and-learn happens, and everyone returns to the workflow they already knew. Adoption metrics look fine. Nothing downstream changed.
Make AIa craftyour teamowns.your team owns.
Hi. I’m David Poyner, a creative technologist and educator. I help creative organizations turn AI into a craft their teams own.
[01]The approach
The problem is rarely the tool.
Why creative AIprograms stall.
Every organization I walk into already has the licenses. What they don’t have is a process that survives contact with a deadline.
Method / 01
We work in cohorts, write the standards down, and measure success by how independently your team operates after I’m gone—not by attendance.
[02]What I run
Four programs. One durable outcome.
Your team gets
better for good.
The work composes: diagnose the system, build fluency on real work, then transfer the practice inside.
02 / 048–16 weeks
Cohort Enablement
A working cohort moves from curiosity to production fluency on live projects from your pipeline—not toy exercises.
- ↳12–30 practitioners working against real briefs
- ↳Live sessions, office hours, and reviewed project work
- ↳Curriculum built on your tools, constraints, and culture
- ↳Internal champions trained to run the next cohort
You get
A fluent team, documented standards, and someone inside who can teach it.
Relevant skills and experience
[Field note 02.1]
Experience is part of the method.
Why it matters ↓Practitioner + educator + systems builder
I make it.
I teach it.
I build the system
behind it.
For more than a decade, my work has sat at the seam between emerging technology and creative practice. That overlap is what makes the capability work useful.
I’ve shipped through four technology transitions.
Interactive advertising. Digital museums. The component-based web. Creative AI. I’ve been inside the work while each new medium moved from spectacle to everyday practice—and learned where adoption actually breaks.
EvidenceCannes Lion · AAM Muse Gold · Webby Honoree
- +Creative AI prototyping
- +Generative image + video
- +Virtual production
- +Creative engineering
[Why this is pertinent]
A tool rollout only changes the work when practitioners trust the person teaching it, the learning survives real deadlines, and the pattern is documented well enough to travel without its inventor. This background covers all three.
[Working fluency]08 capabilities
01Creative AI workflows02Agentic prototyping03Design-to-code04Generative image + video05Virtual production06Curriculum + cohorts07Responsible adoption08Creative systems
[03]How it moves
The engagement is a transfer.
From outside
expertise to
internal craft.
The engagement ends when your team no longer needs me. That is the design constraint—not a concession.
01Listen
See the real workflow.
I sit with the people doing the work, trace the pipeline as it exists, and find the seams where AI can create leverage without breaking trust.
OutputWorkflow map + capability baseline
02Build
Design around live work.
We build the program against current briefs, actual deadlines, your approved stack, and the standards your teams already answer to.
OutputCustom curriculum + working patterns
03Practice
Make fluency visible.
Cohorts work in the open. Feedback is specific, review criteria are shared, and each round turns individual tricks into repeatable craft.
OutputReviewed work + documented standards
04Transfer
Make my role obsolete.
Internal champions run the pattern themselves. The engagement closes when the knowledge has moved out of my head and into your organization.
OutputChampion handoff + next-cohort plan
[04]Ways to work
Choose the depth, not a pile of days.
Find the signal.
Build the change.
[01]
Diagnostic & Keynote
A talk that shifts the room, or a focused assessment that tells leadership where they actually stand. The best front door.
- +Keynote or all-hands
- +Full-day team intensive
- +Or a 2–3 week capability audit
- +Written assessment and roadmap
- +Travel and per diem additional
[02]Core offering
Capability Program
Diagnose, design the curriculum, run the cohort on live work, and transfer the program to internal champions.
- +8–16 weeks, cohort of 12–30
- +Curriculum built on your briefs and stack
- +Documented standards and templates
- +Executive readouts at midpoint and close
- +Milestone billing, 40% to start
[03]
Embedded Practice Lead
Fractional head of creative technology for organizations that need the function in the room while the new process sets.
- +4–6 days of senior time per month
- +Standing weekly working session
- +Async review of creative and technical work
- +Hiring, role-design, and ad-hoc decision support
- +Three-month minimum
- +Quarterly renewal
[05]Why this works
Three credentials. One practice.
The program
needs all three.
01
Educator
My students are working in the field.
Assistant Professor at NYU Tisch since 2020 and co-developer of its inaugural virtual production MPS curriculum. Graduates now work at top virtual production studios, major networks, the United Nations, and leading advertising agencies. Placement is proof that people became capable.
02
Cross-industry
The fourth transition, not the first.
Advertising went interactive. Museums went digital. The web went component-based. I shipped through all three—and now this one. Pattern recognition across transitions is the advantage.
03
Enterprise + studio
It survives organizations that resist it.
Frameworks, pipelines, and standards installed inside global companies and production studios with real inertia, politics, and deadlines. Enthusiastic teams prove little. Skeptical ones do.
[Selected recognition]04
Cannes Lion BronzeCoke Zero Drinkable Billboard
AAM Muse GoldIllinois Holocaust Museum
Communicating the Arts Grand PrixThe MAX
Webby HonoreeMad Libs Mobile
[Selected press]04
CBS · 60 MinutesAI preserves Holocaust survivor conversations
The New York TimesA reimagined Spy Museum
Animation MagazineFive popular virtual production classes
AdweekCoke Zero’s Drinkable Billboard
[06 / Start here]Brooklyn / Remote worldwide
Where is your team stuck?
Let’s make
change stick.
Tell me what your organization is trying to do with AI and where it keeps stalling. I’ll give you a straight answer on fit and what a program could look like.
Start a conversation