Intermediate Level · Cohort 1 · Sold Out
Learn to build reusable creative tools shaped by cinematography, storytelling and the image and video models you already use.
An intermediate level workshop, four sessions, for professional editors and filmmakers. You should already be making things with AI image or video tools, something like Midjourney, Magnific or Seedance. You don't need to be an expert, and you don't need any programming background or prior experience with Claude. Leave with working Skills you built yourself, and the method to keep building more.
Deep Prompting
A deep prompter is a custom Claude Skill built around the knowledge that shapes a good image: cinematography, storytelling and the requirements of the model. Give it a simple idea, and it develops that idea into a complete prompt with shot design, camera language, beats and timing, ready to run.
Its value is control. Your decisions about story, lensing, movement, pacing and continuity become part of a reusable system that you can inspect, edit and refine. The Skill changes when you change it, allowing your process to become more precise and consistent over time while remaining in your hands.
In this workshop, you will build two working deep prompters: one for image prompts across Midjourney, Nano Banana and ChatGPT, and one designed for Seedance video. You will also learn the method for building and adapting new Skills as models and workflows evolve.
What You Leave With
By the end of four sessions you'll have three working tools: something useful you built yourself, a universal image deep prompter and a Seedance video deep prompter. You'll also understand the method behind them, so you can keep building.
Something You Built Yourself
Your first working tool made with Claude, so the technical side stops being intimidating and starts being yours.
A Universal Image Deep Prompter
Develop a rough concept into a detailed image prompt shaped by cinematography and visual storytelling. Works across Nano Banana, Midjourney, and ChatGPT image generation, so you're never tied to one platform.
A Deep Prompter for Seedance Video
Feed it a simple idea, get back a detailed multi-shot video prompt with structured shots, actions, and timed beats.
The Method to Keep Going
Every model is different. You leave knowing how to build the next Skill yourself, for Kling, Veo, LTX, and whatever ships after them.
Live Workshop Format
David leads four live Zoom sessions with structured projects in between. The cohort is capped at twelve, which leaves room for questions, discussion and hands-on guidance.
Sessions are held Thursday evenings, 8:00 to 10:00 PM Eastern / 5:00 to 7:00 PM Pacific.
From David's Workflow
David works with Claude as a creative partner, from concept through generation. These frames from his portfolio reflect the filmmaking judgment behind the Skills and workflows this workshop teaches you to build.
Your Instructor
AI Artist at Wieden+Kennedy · Director · DP
If you've spent any time in Future Film, you already know David. He's one of this community's biggest contributors and a leader in the room: a fixture of the Tuesday sessions, a frequent host of the after-party, and the person our members go to with their hardest AI questions.
David's career reads like preparation for this exact moment. Long before AI pulled technology and filmmaking together, he had gone deep in both: a decade consulting on enterprise systems for IBM, Lockheed Martin, PricewaterhouseCoopers and J.P. Morgan, then twenty years in production, seventeen of them as a union DIT, working with the biggest brands, agencies and studios on LED volumes, high-speed cameras, motion control and motion capture. And he has spent years teaching, which is its own craft: breaking down difficult concepts, making unfamiliar tools approachable, and helping people actually use what they learn. That combination of technical depth, filmmaking experience and teaching ability is rare. It's why David has become such a trusted voice in Future Film and the broader AI film community.
Today he's an AI artist at Wieden+Kennedy on major national campaigns, and the agency's Google Gemini tech liaison: a role that exists because Google required someone on staff who understood its entire AI suite. When W+K went looking for that person, they hired David. He was early to generative filmmaking, has become a go-to voice in the space, and coaches creative directors and speaks at agencies on exactly the ground this workshop covers. In his own words: AI is just the newest tool in his kit.
Frames from David's 2026 reel.
Why This Matters Now
Creative leverage comes from turning taste, judgment and production knowledge into systems you can use and refine.
A prompt gets you one image. A Skill gets you a repeatable way of working: your taste, your standards and your cinematography knowledge, encoded into a tool you can use and refine.
What makes editors and filmmakers uniquely positioned for this is everything they already know. Story instinct, pacing, coverage, continuity. Claude turns that knowledge into leverage. This workshop teaches you how.
And because you build the Skills yourself, you leave with more than tools. You leave knowing how to make the next one.
Curriculum
Schedule
September 17 to October 8, 2026
Thursday evenings via Zoom · 2 hours per session
8 PM Eastern / 5 PM Pacific
David walks the room through getting Claude Code installed and configured properly: terminal, permissions, file structure, and the pieces that stop most people before they start. You'll learn how the setup works and why, not just click through it, so the environment you build on for the rest of the workshop is a solid one.
Keep working through any remaining setup on your own machine, with support in the cohort Discord channel.
The foundation: the difference between Claude Chat, Cowork, and Claude Code, and when to use each. How Projects and Skills fit together. Best practices for organizing your work. How credits and usage actually work and how to keep costs under control. How to start treating Claude as a repeatable creative system.
Build something functional of your own with Claude: a Skill, a small creative tool, a utility, a simple game. The point is that you make something that works, yourself.
Build a reusable deep prompter for image generation that incorporates cinematography and visual storytelling. Learn to make detailed changes, improve control and consistency, and work across image models instead of being tied to one platform.
Take a script, pitch, deck, or brief and use Claude to turn it into something you could present to a client: breakdowns, story beats, visual references, character and location material. Use your own project, or we'll provide one.
Apply the same approach to video, focused on Seedance and its multi-shot capabilities. Structure shots, actions, and timed beats. Turn a simple idea into a detailed, controlled video prompt. Leave with a working Seedance deep prompter and the foundation to build model-specific Skills for Kling, Veo, LTX, and beyond.
Who This Is For
This is an intermediate workshop. You don't need to be an expert, and you don't need any programming background. You should have some hands-on experience with AI image or video tools, things like Midjourney, Magnific or Seedance, and be ready to build a stronger system around the work you're already doing. Brand new to generative tools? Join our free Future Film community. We have tons of resources, and workshops for all levels coming soon.
What You'll Need
Workshop Details
Starts
Sept 17
2026, Pilot cohort
Format
Live Zoom
4 sessions · 2 hours each
Thursdays, 8 PM Eastern / 5 PM Pacific
Tuition
$600
Pilot cohort pricing · 12 seats max
Cohort 1 is full — join the waitlist below
What's Included
Future Film
Future Film is an ongoing community of editors, filmmakers, and AI artists exploring emerging storytelling and production workflows together, live, every Tuesday night. Workshop students become part of that community and stay connected to a network of people doing the same work.
Visit futurefilm.club ↗Questions
This is an intermediate level workshop. You don't need to be an expert, we'd just ask that you not be a beginner. You should have some hands-on experience with AI image or video tools, things like Midjourney, Magnific or Seedance. You don't need any experience with Claude, Claude Code, or agentic workflows. That's what the workshop is for.
No. If you've never opened a terminal, Session 1 exists for exactly that. No programming background is required or expected.
You're enrolled as soon as payment is complete. A welcome email follows with your seat confirmation, then onboarding details: calendar links, your invite to the private Discord channel, and a short setup survey before Session One.
Every session is recorded and available to participants, and the Discord channel keeps the conversation going between sessions.
Cohort 1 filled on Tuesday, September 15. Join the waitlist for Cohort 2.
Cohort 1 · Sold Out · September 2026
The first cohort is at capacity. Join the waitlist and we will reach out when the next cohort opens or a spot becomes available.