11 Dollar Bill Overhauls Ad Production With Concurrent Generative AI

Blow Up Your Old Workflow, Find Creative Flow With 11 Dollar Bill and Agent11 — Ai4

Ai4

Restructuring linear post-production into concurrent generative AI pipelines eliminates turnaround bottlenecks and empowers multidisciplinary creative generalists while protecting human talent compensation.

AI-assisted summary of Blow Up Your Old Workflow, Find Creative Flow With 11 Dollar Bill and Agent11 — original source. Check the original for full context.

Source published: 2025-08-27

Detailed notes

## Background Published by the Ai4 channel on YouTube on August 27, 2025, this presentation features Christian Robins, Chief Creative Officer and co-founder of post-production studio 11 Dollar Bill and GenAI consultancy Agent11. According to the description, the session explores how multidisciplinary creatives can integrate generative AI into practical workflows to remove production bottlenecks, accelerate iteration, and amplify human storytelling in modern advertising.

## TL;DR Christian Robins makes the case for replacing rigid, linear post-production pipelines with parallel, concurrent workflows empowered by generative AI. Walking through a phased adoption framework, he demonstrates real-world case studies including concept ideation with Midjourney and Runway, instant music generation in Suno, and an ethically governed ElevenLabs voice model used to lock Domino's commercial pacing prior to talent recording. Robins also details complex visual applications, such as iPhone-based Luma AI 3D scans for JD Sports and synthetic environment compositing for Florastor, illustrating how AI can eliminate operational friction and empower agile 'creative generalists' without compromising human artistry.

## Atmosphere & style The presentation carries an energetic, pragmatic, and refreshingly candid tone, delivered with the grounded authority of a veteran editor and creative director who demystifies industry hype. Robins balances technical pipeline breakdowns with quick humor and relatable war stories from commercial suites, maintaining a reassuring emphasis on craft preservation and ethical talent governance. Watching the session feels like a brisk, backstage tour of a modern post house, culminating in a playful synthetic avatar sign-off that earns enthusiastic praise from the conference host.

## Key takeaways - [00:04](https://www.youtube.com/watch?v=zuNWlE82ujI&t=4s) Generative AI does not replace artistic craft, but removes procedural friction so creatives can stay in creative flow and focus on high-value experimentation. - [02:46](https://www.youtube.com/watch?v=zuNWlE82ujI&t=166s) Traditional post-production follows a slow, rigid, linear pipeline where problems surface late, driving up costs and requiring expensive pickup shoots. - [02:46](https://www.youtube.com/watch?v=zuNWlE82ujI&t=166s) 11 Dollar Bill replaced the linear assembly line with a concurrent workflow where editorial, motion graphics, temp audio, and visual effects happen simultaneously in-suite. - [04:33](https://www.youtube.com/watch?v=zuNWlE82ujI&t=273s) Treating generative AI as a modular 'menu' of assists allows teams to adapt tooling per project while transforming vendor relationships into transparent client workshops. - [04:33](https://www.youtube.com/watch?v=zuNWlE82ujI&t=273s) Contractual guardrails and explicit consent around talent likeness, voice models, and data sourcing are vital to building client trust in GenAI pipelines. - [07:17](https://www.youtube.com/watch?v=zuNWlE82ujI&t=437s) Small-scale adoption starts with visual ideation and storyboarding using tools like Midjourney and Runway, eliminating blank page syndrome before physical crews are hired. - [09:04](https://www.youtube.com/watch?v=zuNWlE82ujI&t=544s) Generative audio tools like Suno provide on-demand music generation directly within the edit suite, avoiding external composition delays for quick-turnaround spots. - [09:04](https://www.youtube.com/watch?v=zuNWlE82ujI&t=544s) Domino's uses an authorized ElevenLabs voice clone ('Robo-Casey') to test script pacing internally, while paying the human voice talent in full to record the final broadcast audio. - [13:27](https://www.youtube.com/watch?v=zuNWlE82ujI&t=807s) Generative video and inpainting tools salvage footage by transforming weather and instantly removing lighting gear and cables without green screens. - [13:27](https://www.youtube.com/watch?v=zuNWlE82ujI&t=807s) Mobile LiDAR scanning with Luma AI enables rapid 3D asset generation and compositing for social campaigns in days at a fraction of traditional production budgets. - [16:35](https://www.youtube.com/watch?v=zuNWlE82ujI&t=995s) Combining multiple tools, such as Adobe Firefly and Runway, allows productions to blend live-action green screen talent into fully synthetic worlds. - [18:08](https://www.youtube.com/watch?v=zuNWlE82ujI&t=1088s) The long-term objective of GenAI adoption is building bespoke studio infrastructure and agentic workflows to tackle challenges standard software cannot solve. - [18:08](https://www.youtube.com/watch?v=zuNWlE82ujI&t=1088s) The rise of the 'creative generalist' empowers multidisciplinary artists to eliminate routine bottlenecks and form agile, highly productive creative teams. - [19:33](https://www.youtube.com/watch?v=zuNWlE82ujI&t=1173s) Creative organizations should approach generative AI adoption systematically across small, medium, and large phases rather than attempting an overnight overhaul.

## Detailed walkthrough ### [00:04](https://www.youtube.com/watch?v=zuNWlE82ujI&t=4s) Introduction and the Quest for Creative Flow

Christian Robins introduces himself as the Chief Creative Officer and co-founder of 11 Dollar Bill, a post-production studio with offices spanning Boulder, Colorado, Chicago, and Raleigh-Durham, North Carolina. Over a career spanning more than 25 years, Robins and his team have created commercials, documentaries, and large-format experiential projects for prominent brands including Domino's, Adidas, Popeyes, JD Sports, Netflix, Manchester United, and even high-profile content for the Las Vegas Sphere. Robins jokes about his long-standing "bad habit of early adopting tech before it's ready," framing his core role as sifting through industry noise and technical friction to curate what actually works inside real-world creative workflows.

The talk centers on the philosophy of "blowing up your old workflow"—interrogating established production habits and reinventing them to attain "creative flow." While acknowledging that many creatives feel deeply threatened by generative AI because their artistic identity is bound up in execution and manual craft, Robins presents an alternative perspective. Generative AI does not replace artistic vision; rather, when implemented pragmatically, it amplifies vision and strips away tedious, repetitive tasks so teams can spend their time experimenting and creating.

- **Studio Locations:** Boulder (CO), Chicago (IL), and Raleigh-Durham (NC). - **Core Philosophy:** "Keep momentum, protect the craft, and then let AI remove the friction so we can all kind of stay in flow."

---

[02:46](https://www.youtube.com/watch?v=zuNWlE82ujI&t=166s) Blowing Up the Linear Pipeline for Concurrent Workflows

Traditional post-production operates in a rigid, sequential assembly line: filming footage on set, passing it to editorial, handing off cuts to motion graphics and design, scoring and mixing audio, moving into finishing and color grading, and finally reaching delivery. Robins explains that this linear model frequently breaks creative flow because problems surface late, making revisions exponentially expensive and forcing teams to burn time and money re-executing concepts that miss the mark.

```mermaid flowchart TD accTitle: Comparison of Traditional Linear Post-Production and 11 Dollar Bill's Concurrent AI Pipeline subgraph Traditional["Old Linear Pipeline (Slow & Rigid)"] direction LR T1["Production / Shoot"] --> T2["Editorial"] T2 --> T3["Graphics / VFX"] T3 --> T4["Audio Mix"] T4 --> T5["Finishing & Delivery"] end

subgraph Modern["Modern Concurrent Pipeline (Rapid Iteration)"] direction TB M0["Client Ideation & Scripting"] subgraph ConcurrentSuite["Live Suite Collaboration"] M1["Editorial"] M2["Motion Graphics"] M3["GenAI Temp Audio / VO"] M4["Rapid Scratch Frames / VFX"] end M0 --> ConcurrentSuite ConcurrentSuite --> M5["Final Polish & Fast Delivery"] end ```

To solve these bottlenecks, 11 Dollar Bill restructured their pipeline into a parallel, concurrent system where departments stop waiting on one another. Motion graphics, editorial, visual effects, temp music generation, and scratch voiceovers now occur simultaneously in the room. This shift allows problems to surface immediately and decisions to be validated with concrete visual evidence early in the schedule. A single creative can now execute multi-disciplinary tasks that previously required siloed teams, allowing small units to operate with exceptional speed and agility.

- **Old Pipeline Flaws:** Slow iteration, linear dependencies, delayed error discovery, high revision costs, and extensive pickup shoots. - **New Workflow Benefits:** Concurrent audio/visual prototyping, early decision-making, reduced pickups, and consolidated creative capabilities.

---

[04:33](https://www.youtube.com/watch?v=zuNWlE82ujI&t=273s) The GenAI Menu and Transparent Client Collaboration

Robins treats generative AI not as a rigid corporate mandate, but as a modular "menu" of assists tailored to the specific needs of each project. Certain jobs might only need a single AI tool for rough scratch audio, while others leverage multiple tools across ideation, visual effects cleanup, and asset generation. By removing procedural friction during pre-production and concept development, client sessions transform from formal vendor transactions into collaborative creative workshops.

Maintaining full transparency is essential to this dynamic. 11 Dollar Bill educates clients on the mechanics of each tool, detailing where data is sourced and establishing clear contractual guardrails around talent likeness, voice consent, and editorial scope. Rather than presenting generic one-page service sheets, the studio performs live demonstrations directly in the editing suite, giving clients direct visibility and comfort with the technology upstream.

- **Collaborative Dynamic:** Shifting the relationship from transactional vendor delivery to "a group of creative people in a room working together." - **Governance & Trust:** Strict contractual consent protocols for voice and likeness, defined guardrails, and transparent sourcing.

---

[07:17](https://www.youtube.com/watch?v=zuNWlE82ujI&t=437s) Small Steps: Breaking Blank Page Syndrome with Visual Ideation

The practical implementation of AI begins with small steps, primarily targeting "blank page syndrome" during early ideation. For a major Dave & Buster's commercial campaign involving store redesigns, 11 Dollar Bill ingested treatments and existing design assets into Midjourney and Runway to generate comprehensive, polished storyboards before hiring crew members or stepping on set. This alignment locked in the director's vision immediately and ensured all stakeholders shared the exact visual expectations upfront.

Beyond conventional storyboarding, generative tools democratize the visual pitch process across all roles. Anyone on the team—from junior coordinators to senior producers—can articulate abstract or surreal concepts by generating direct visual references, such as visualizing Pepsi raining from the sky or transforming a lemon into a bird to guide the VFX department.

- **Dave & Buster's Case Study:** Used Midjourney and Runway to generate comprehensive visual storyboards, establishing creative alignment before hiring physical production crews. - **Democratizing Pitching:** Enables non-designers and producers to bypass technical barriers and generate clear visual targets for VFX teams.

---

[09:04](https://www.youtube.com/watch?v=zuNWlE82ujI&t=544s) Generative Audio: Suno Jingles and Domino's "Robo-Casey"

Generative audio provides significant time savings during the edit. Robins demonstrates this using Suno, playing a snippet of a 1990s-style West Coast hip-hop beat generated prior to the presentation, and highlights a commercial execution for Ocean Spray. Tasked with crafting a rapid social media video around a quirky client concept, the editor quickly generated an orchestral choral track singing the product tagline directly within the suite, avoiding external composition delays.

``` [Domino's Editorial Workflow with Robo-Casey]

1. Script Draft & Footage Ingest │ ▼ 2. ElevenLabs "Robo-Casey" Generation (Scratch Audio) ├── Edit cuts to temp audio ├── Check cadence, timing, and line length └── Rapidly rewrite & re-generate script variants in-suite │ ▼ 3. Final Script Locked & Sent to Talent │ ▼ 4. Casey Records Final Voiceover (Full Talent Compensation) │ ▼ 5. Final Mix Placed into Master Cut (Consumer Delivery) ```

The studio also relies heavily on ElevenLabs for voiceover workflows, particularly on high-volume broadcast and social content for Domino's. Domino's relies on a single voice talent named Casey across all advertising. Because scripts frequently require pacing and wording adjustments once laid against raw picture, the client created a custom ElevenLabs model trained with Casey's explicit consent. Editors use "Robo-Casey" to test rewrites and lock comedic timing in real time. Once pacing is perfected, the clean final script is sent to Casey for the actual master recording—ensuring the scratch model never reaches consumers, talent gets paid full compensation, and editors save hours of idle turnaround time.

- **Ocean Spray Social Spot:** Generated a custom choral jingle via Suno on demand inside the edit suite. - **Domino's "Robo-Casey":** Trained an authorized ElevenLabs voice clone to test script revisions and pacing internally; Casey records all final broadcast audio. - **Human-in-the-Loop Safeguard:** "Casey gets paid the same. Makes her life a hell of a lot easier. It makes our life a hell of a lot easier. Why wouldn't you be using these tools?"

---

[13:27](https://www.youtube.com/watch?v=zuNWlE82ujI&t=807s) Generative Video, Fixes, and 3D Scanning with Luma AI

Generative video tools eliminate the time sink of scrolling through generic stock websites by allowing artists to generate bespoke assets or extend existing live-action footage. For Vail Resorts, 11 Dollar Bill took footage captured during overcast, dreary conditions and transformed the environment into a bright, bluebird sunny day, salvaging massive production investments. Generative tools also streamline routine cleanup work: when zooming out on high-resolution 6K or 8K footage exposes lighting equipment, C-stands, or loose cables, editors can eliminate them on the spot rather than pushing cleanups to downstream VFX schedules.

For a fast-turnaround social media activation for JD Sports promoting a The North Face jacket, the team stuffed the jacket with Amazon shipping boxes to resemble a gargoyle perched atop a skyscraper. They captured a LiDAR scan using an iPhone running Luma AI inside their producer's office during winter, generated a detailed 3D model, and tracked the rendered jacket onto the crown of the Empire State Building. The stunt generated millions of organic views within days at a fraction of standard production timelines.

- **Vail Resorts Weather Transformation:** Converted overcast ski footage into sunny, bluebird mountain scenes. - **Set & Plate Cleanup:** Instant removal of C-stands, lighting gear, cables, and background elements without green screens. - **JD Sports & Luma AI:** Mobile LiDAR scan of a padded The North Face jacket composited onto the Empire State Building, produced in days and generating millions of views. - **Industry Reference:** Robins highlights Luma AI co-founder Amit Jain's fireside session as essential viewing for generative 3D pipelines.

---

[16:35](https://www.youtube.com/watch?v=zuNWlE82ujI&t=995s) Taking Bigger Steps: Multi-Tool Integration for Florastor

Moving beyond isolated tasks, 11 Dollar Bill began combining multiple generative AI platforms into unified commercial pipelines. Working alongside the probiotic brand Florastor and their agency, the studio aimed to establish a first-to-market advantage by using generative visuals in their sector. The team combined Adobe Firefly and Runway during visual development to create unique concept designs, which were then executed via a traditional green screen shoot to blend live-action performance with synthetic environments.

- **Florastor Campaign Goal:** Achieve first-to-market generative visual advertising in the probiotic health sector. - **Multi-Tool Pipeline:** Integrated Adobe Firefly and Runway for concept development and environment styling prior to green screen stage capture.

[17:10](https://www.youtube.com/watch?v=zuNWlE82ujI&t=1030s) Synthetic Worlds: The Florastor Commercial and the Generative Workflow Loop

Christian Robins showcases a real-world production example demonstrating the fusion of live-action talent with fully generative environments. Using image-to-video generative tools, the team established a custom, proprietary visual aesthetic for a commercial where live-action actors are composited directly into completely synthetic surroundings. Presenting the finished spot, the voiceover highlights the concept: *"It's another superb day in Gut City thanks to the flora restorers working hard to flush out the bad and boost your own good bacteria. Thanks to them, your gut is more balanced and resilient with Florastor probiotics. It's all systems go."* Robins reveals the underlying reality of the production: "The only thing real on that spot are the people. Everything else is for fun."

```mermaid flowchart LR accTitle: Generative Production Workflow Loop subgraph Pre["Creative Ideation"] A["Concept & Style"] --> B["AI Environment Gen"] end subgraph Prod["Production & Composite"] C["Live Talent Shoot"] --> D["Image-to-Video & Composite"] end subgraph Post["Final Delivery"] D --> E["Finished Commercial Spot"] E -.->|"Refine Loop"| A end ```

This project exemplifies the iterative generative workflow loop in full swing. Rather than forcing every production through a rigid, traditional pipeline, the generative approach allows artists to touch and shape every phase of the sequence while continuously refining the process into an bespoke pipeline tailored specifically to the project's creative needs.

- **Real vs. Synthetic Breakdown**: Live actors were the sole practical element filmed on set; all complex background environments, architecture, and atmospheric effects were created through generative AI workflows. - **Workflow Iteration**: The pipeline is designed as an adaptable loop where prompt generation, image synthesis, and video extension feed back into compositing and finishing.

---

[18:08](https://www.youtube.com/watch?v=zuNWlE82ujI&t=1088s) Creative Infrastructure and the Rise of the "Creative Generalist"

Robins emphasizes that the endgame of adopting AI is not just using disparate apps, but constructing comprehensive, custom AI workflows for unique creative challenges that traditional post-production tools cannot address. By training agents infused with their own studio's creative identities, 11 Dollar Bill and Agent11 build custom creative infrastructure. This technological scaffolding is assembled by experienced craftspeople who understand movement, timing, and storytelling, allowing them to inject their inherently human creative instincts into the pipeline without compromising artistic standards.

The primary objective of these integrated workflows is eliminating operational friction to create an environment where creatives never have to compromise or tell a client "no"—they can simply execute. As practitioners experiment across varied AI tools, they evolve beyond siloed roles into multidisciplinary talent. Robins highlights the concept of the "creative generalist," an idea he attributes to his friend Brent Peterson at Signal & Cipher and includes in every presentation:

> "Anybody who has a strength and has all these ancillary strengths now can take these tools, eliminate the nightmares out of their day, and then surround themselves with people who are also interested in all different types of content and tools... and then you have these small teams that are totally unstoppable now."

- **Agentic Workflows**: Custom AI agents trained on specific studio identities tackle complex creative barriers beyond the reach of standard software. - **Human-in-the-Loop Infrastructure**: Designed to preserve and channel human craft rather than automate it away. - **The Creative Generalist Model**: Empowering artists with deep core capabilities and diverse secondary interests to form agile, hyper-productive creative units.

---

[19:33](https://www.youtube.com/watch?v=zuNWlE82ujI&t=1173s) Phasing Adoption and Closing at Ai4

Robins concludes by outlining a manageable roadmap for studios and agencies looking to modernize their creative operations. Rather than overhauling an entire studio overnight, teams should implement shifts systematically—starting small, scaling to medium optimizations, and finally tackling large-scale pipeline transformations one workflow shift at a time.

Wrapping up the session, Robins triggers a synthetic avatar video that delivers a playful farewell to the audience: *"Hey there, Ai4. Thanks for listening to me ramble today about generative AI. This is pretty neat."* Agreeing with his digital counterpart—"And she's right. This is pretty neat"—Robins thanks the audience and wishes everyone a great conference in Las Vegas. The session moderator takes the stage to close out the segment with high praise, declaring to the room, *"Everybody give it up for Christian Robins. That was easily my favorite presentation all day. I must say that was awesome."*

## Notable quotes > [00:04](https://www.youtube.com/watch?v=zuNWlE82ujI&t=4s) bad habit of early adopting tech before it's ready — Christian Robins

> [00:04](https://www.youtube.com/watch?v=zuNWlE82ujI&t=4s) Keep momentum, protect the craft, and then let AI remove the friction so we can all kind of stay in flow. — Christian Robins

> [04:33](https://www.youtube.com/watch?v=zuNWlE82ujI&t=273s) a group of creative people in a room working together. — Christian Robins

> [09:04](https://www.youtube.com/watch?v=zuNWlE82ujI&t=544s) Casey gets paid the same. Makes her life a hell of a lot easier. It makes our life a hell of a lot easier. Why wouldn't you be using these tools? — Christian Robins

> [17:10](https://www.youtube.com/watch?v=zuNWlE82ujI&t=1030s) It's another superb day in Gut City thanks to the flora restorers working hard to flush out the bad and boost your own good bacteria. Thanks to them, your gut is more balanced and resilient with Florastor probiotics. It's all systems go. — Commercial Voiceover

> [17:10](https://www.youtube.com/watch?v=zuNWlE82ujI&t=1030s) The only thing real on that spot are the people. Everything else is for fun. — Christian Robins

> [18:08](https://www.youtube.com/watch?v=zuNWlE82ujI&t=1088s) Anybody who has a strength and has all these ancillary strengths now can take these tools, eliminate the nightmares out of their day, and then surround themselves with people who are also interested in all different types of content and tools... and then you have these small teams that are totally unstoppable now. — Christian Robins

> [19:33](https://www.youtube.com/watch?v=zuNWlE82ujI&t=1173s) Hey there, Ai4. Thanks for listening to me ramble today about generative AI. This is pretty neat. — Synthetic Avatar

> [19:33](https://www.youtube.com/watch?v=zuNWlE82ujI&t=1173s) And she's right. This is pretty neat — Christian Robins

> [19:33](https://www.youtube.com/watch?v=zuNWlE82ujI&t=1173s) Everybody give it up for Christian Robins. That was easily my favorite presentation all day. I must say that was awesome. — Session Moderator

## Mentioned Christian Robins, 11 Dollar Bill, Agent11, Ai4, Domino's, Adidas, Popeyes, JD Sports, Netflix, Manchester United, Las Vegas Sphere, Dave & Buster's, Midjourney, Runway, Suno, ElevenLabs, Casey, Ocean Spray, Vail Resorts, The North Face, Luma AI, Amit Jain, Empire State Building, Florastor, Adobe Firefly, Brent Peterson, Signal & Cipher

## Video description Christian Robins, CCO/Co-Founder, 11 Dollar Bill and Agent11

Abstracts

Generative AI is transforming advertising, empowering AI-Enhanced Creatives to work faster, iterate seamlessly, and push the boundaries of storytelling. In this session, Christian Robins, multiple Emmy Award-winning creative and co-founder of post-production studio 11 Dollar Bill and GenAI consultancy Agent11, shares his practical, workflow-driven approach to integrating generative AI into real-world creative processes.

Christian will break down small, medium, and big steps for AI adoption - demonstrating how tools like AI-powered ideation, generative video, and synthetic media are revolutionizing creative workflows. Learn how AI can enhance collaboration, remove bottlenecks, and accelerate content creation pragmatically - not replacing human creativity, but amplifying its potential.

This session explores how AI-Enhanced Creatives - multidisciplinary experts who merge AI tools with traditional creative skills - are leading the next wave of advertising innovation. Walk away with actionable strategies to harness generative AI, refine your creative process, and unlock new possibilities in modern advertising.

[Source](https://www.youtube.com/watch?v=zuNWlE82ujI)

中文简报

11 Dollar Bill利用并发生成式AI重塑广告制作将线性后期制作重构为并发生成式AI管线，既消除了交付周期瓶颈，赋能跨领域的创意全才，又保障了人类人才的报酬权益。 Christian Robins 主张用生成式AI赋能的并行并发工作流取代僵化、线性的后期制作管线。通过梳理分阶段落地框架，他展示了多个真实案例，包括使用 Midjourney 和 Runway 进行概念构思、在 Suno 中即时生成音乐，以及在演职人员录音前使用受伦理规范监管的 ElevenLabs 语音模型来敲定 Domino's 广告的节奏。Robins 还详述了复杂的视觉应用，例如为 JD Sports 进行基于 iPhone 的 Luma AI 3D 扫描，以及为 Florastor 进行合成环境合成，展示了 AI 如何在不牺牲人类艺术创造力的前提下消除操作阻力，赋能敏捷的“创意全才”。 11 Dollar Bill 摒弃了传统缓慢僵化的线性后期制作流水线，转向剪辑、动态设计、临时音频与视觉特效在套间内同步推进的并行工作流，从而及早发现问题并大幅减少昂贵的补拍成本。 (2:46) 团队将生成式 AI 视为模块化的工具“菜单”，根据项目需求灵活调用，并通过针对肖像权、声音模型与数据来源设立严格的合同规范与授权机制，与客户建立起透明可信的协作关系。 (4:33) 在商业实战中，Domino's 采用获授权训练的 ElevenLabs 声音模型“Robo-Casey”在套间内实时测试脚本节奏与喜剧包袱，最终广播录音仍全额向真人配音演员支付酬劳，兼顾了制作效率与从业者权益。 (9:04) 生成式 AI 的深入应用催生了“创意通才”（creative generalist），资深创作者将自身审美与自定义 Agent 基础设施深度融合，构建出极具敏捷性与高产出效能的小型创意团队。 (18:08)

Reading summary

Christian Robins, Chief Creative Officer of post-production studio 11 Dollar Bill, argues that integrating generative AI into production pipelines does not replace human craft, but removes operational friction by breaking down rigid, linear post-production workflows into agile, concurrent processes. Replacing the Linear Post-Production Assembly Line (2:46) Traditional post-production follows a sequential chain running from shooting to editorial, visual effects, audio mixing, and final color grading. Robins explains that this structure delays problem discovery, making revisions late in the process expensive and frequently requiring pickup shoots. To resolve these bottlenecks, 11 Dollar Bill restructured its workflow into a parallel pipeline where editorial, scratch audio, visual effects, and motion graphics are prototyped concurrently in the suite, allowing teams to validate decisions earlier and reduce turnaround times. Early Ideation and Visual Pitching (7:17) Adopting generative tools begins with small implementations to tackle blank-page syndrome during pre-production. In a campaign for Dave & Buster's, the studio fed treatments and existing design assets into Midjourney and Runway to produce finished visual storyboards before hiring crews or arriving on set, securing creative alignment early. Robins notes that these generative tools democratize the pitching process across departments, enabling non-designers and producers to generate concrete visual references for abstract concepts, such as visualizing Pepsi falling as rain or turning a lemon into a bird to instruct visual effects artists. Modular Audio, Video Fixes, and Voice Safeguards (9:04) 11 Dollar Bill applies generative AI as a project-specific menu of assists rather than a mandatory mandate. In audio workflows, the studio used Suno to produce an on-demand choral jingle for an Ocean Spray social spot. For Domino's, the team addressed script-timing bottlenecks by building an authorized ElevenLabs voice clone of their primary voice talent, Casey. Editors use "Robo-Casey" internally to refine script cadence and comedic timing, while Casey is paid her standard rate to record the approved final script, ensuring synthetic audio never reaches consumer airwaves. On the visual side, generative tools replace stock library searches and salvage costly shoots. For Vail Resorts, 11 Dollar Bill transformed overcast mountain footage into bright, sunny scenes and routinely removes unwanted gear like C-stands and cables directly in the editing room. For JD Sports, the team scanned a stuffed jacket inside an office using Luma AI's mobile LiDAR, converted it into a 3D asset, and composited it atop the Empire State Building, generating millions of social impressions within days. Integrated Multi-Tool Pipelines and Creative Generalists (16:35) Moving beyond isolated applications, the studio builds end-to-end multi-platform pipelines. In a campaign for the probiotic brand Florastor, live-action actors filmed on a green screen stage were composited into fully synthetic environments generated via Adobe Firefly and Runway. The final commercial spot, set in a conceptual "Gut City," contained live actors as the only practical physical element, with all backdrops and architectural details produced through an iterative generative loop. Robins highlights that this infrastructure supports the rise of the "creative generalist"—a concept he credits to Brent Peterson of Signal & Cipher. Rather than relying on isolated specialists passing tasks sequentially down an assembly line, multidisciplinary artists equipped with custom agentic workflows can manage multiple stages of production, allowing smaller teams to solve complex creative problems rapidly without lowering craft standards. Robins advises studios and agencies to phase in generative tools systematically—progressing from small procedural aids to medium optimizations before overhauling entire pipelines—to protect human artistic intuition while eliminating operational drag.

精炼阅读摘要

11 Dollar Bill联合创始人兼首席创意官Christian Robins指出，生成式人工智能并非取代创作者的手工技艺与艺术愿景，而是通过解构传统影视后期制作的刚性线性流程，释放团队的创作心流。他主张在坚守肖像与声音确权等伦理底线的前提下，将AI工具模块化融入制作各环节，以推动“创意通才”主导的高效敏捷团队模式。重构制作管线：从线性交付转向并发协作 (2:46) 传统影视后期制作遵循拍摄、剪辑、动效与视效、音频混音及最终调色交付的严格单向流水线。Robins指出，这种线性机制极易打断创作者的心流状态，因为制作偏差往往在项目尾声才暴露，导致后期返工与追加补拍的成本呈指数级增长。为此，11 Dollar Bill将工作流重组为多部门协同的并行网络，在剪辑套件内同步推进画面剪辑、动态图形、临时AI音频生成与视觉特效概念占位。这种并发工作模式让制作团队能够在项目初期即时发现问题，并通过具体的具象化视听证据验证决策。由于AI工具降低了技术执行门槛，单一创意人员可跨界承担原先被工种壁垒分割的多元任务，使小型创意单元能以极高的迭代效率推进复杂项目。模块化工具箱与声音授权治理机制 (4:33) 该工作室将生成式AI视为可根据具体项目灵活调取的模块化“菜单”，而非一刀切的硬性指标。为打破传统的甲乙方外包壁垒，团队在机房内直接向客户进行现场操作演示，透明解释各工具的数据来源与底层逻辑，使客户从被动的审片方转变为前期的共创伙伴。在版权与合规方面，工作室建立了严格的合约授权护栏。以达美乐（Domino's）的广告项目为例，品牌长期合作配音演员Casey授权其声音数据训练了专用的ElevenLabs声音模型“Robo-Casey”。剪辑师在机房内利用该AI模型即时测试新文案的语速、停顿与喜剧节奏；待成片脚本彻底定稿后，再交由Casey本人进行正式母带录制并支付全额报酬。这种“人在回路”（Human-in-the-Loop）机制既规避了未授权合成音频面向公众播出的风险，又大幅缩短了反复沟通录音的周转时间。视觉概念推演、实拍修复与轻量3D (7:17) 在前期创意推演中，生成式视觉工具被用于破解“白纸综合征”。例如在Dave & Buster's的店面升级商业企划中，团队将设计方案输入Midjourney与Runway，在正式聘用剧组与实地开机前便生成了高完成度的视觉分镜，统一了客户与导演的预期。此类工具还消除了非美术岗位的沟通阻碍，使初级制片人亦能迅速生成如“可乐从天而降”或“柠檬变飞鸟”等具象参考画面。在实拍后期阶段，AI工具显著降低了寻找素材与繁琐手工擦除的损耗。在Vail Resorts的滑雪度假村项目中，阴沉多云的实拍素材被算法转化为晴空万里的人造实景；在常规6K至8K素材镜头缩放时，剪辑师可当场清除画框边缘穿帮的C型架与灯光线缆。在JD Sports针对北面（The North Face）羽绒服的社媒营销中，团队仅用iPhone配合Luma AI对塞满纸箱的羽绒服进行激光雷达（LiDAR）扫描建模，随后合成至帝国大厦顶端，以极低预算在数天内实现了数百万次有机传播。虚实融合实践：复合工具与生成式环境 (17:10) 在益生菌品牌Florastor的商业广告战役中，11 Dollar Bill尝试了多平台复合的生成式商业片制作闭环。团队在前期视觉开发阶段打通Adobe Firefly与Runway，推演独特的概念风格，并在绿幕摄影棚内完成真实演员的实拍，旨在该健康细分赛道建立首发视觉优势。在最终成片中，除演员本人的肢体表演外，所有微观建筑、肠道菌群与大气光影等复杂环境均由图像转视频（Image-to-Video）AI工具生成合成。该案例验证了从提示词构思、风格化图像合成到视频拓展与最终合成的动态迭代闭环，展示了在受控商业制作中完全由AI生成沉浸式虚实融合场景的可行性。定制智能体与“创意通才”敏捷团队 (18:08) Robins强调，生成式AI的落地不应停留在零散使用第三方软件，而应由深谙镜头运动、叙事与剪辑节奏的专业影视人主导，训练内嵌工作室自身创意特质的定制AI智能体（如Agent11），搭建契合自身审美标准的创意基础设施。这一架构重塑了创意组织的形态。借用Brent Peterson提出的“创意通才”（Creative Generalist）概念，Robins认为掌握一项核心技能并涉猎多项边缘能力的从业者，借助AI工具消除琐碎的技术摩擦后，能够打破传统部门壁垒。由这类通才组建的小型敏捷团队，无需以牺牲创意标准为代价向技术限制妥协，具备极强的执行力与生产效率。 Robins总结认为，机构推进AI转型应遵循从小步尝试、局部优化到系统重塑的分阶段路线，以保护核心工艺为前提化解技术摩擦。需要指出的是，本案讨论的经验主要基于商业广告、社媒推广等高周转率影视制作，生成式环境在超长叙事体裁及复杂物理交互中的稳定性，以及跨域声音与视觉版权在更大行业范围内的法律界定，仍有待行业长期实践检验。

Detailed notes — zh

## Background 该视频由 Ai4 于 2025 年 8 月 27 日发布在 YouTube 平台上。内容记录了多次荣获艾美奖的后期制作工作室 11 Dollar Bill 兼 GenAI 咨询机构 Agent11 联合创始人兼首席创意官 Christian Robins 的主题演讲。根据官方简介，演讲聚焦于如何以实用且以工作流为核心的方式将生成式 AI 融入真实创意流程，阐述跨学科的“AI 赋能型创意人”如何打破传统制作瓶颈并放大人类创造力。 ## TL;DR Christian Robins 主张彻底颠覆传统后期制作中僵化的单向线性流程，转向剪辑、视觉特效与音频实时并行的全新 AI 协作系统。通过达美乐（Domino's）、优鲜沛（Ocean Spray）、JD Sports 及 Florastor 等大量商业实战案例，他展示了如何运用 ElevenLabs、Suno、Runway 和 Luma 3D 等模块化工具实现即时概念分镜、内部占位配音与快速三维资产生成。他强调，生成式 AI 的核心价值在于消除机械性的流程摩擦，保护艺术家的“创意心流”，同时将甲乙方的买卖关系重塑为透明共创模式。最终，兼具专业深度与多维技能的“创意通才”将借助定制 AI 基础设施，组成势不可挡的高效敏捷小团队。 ## Atmosphere & style 整个演讲散发着资深影视制作人特有的务实与激情，语言风趣幽默、节奏明快，毫无晦涩的技术说教。Christian 带着对影像和音乐创作的纯粹热爱侃侃而谈，从即兴生成西海岸说唱到用 iPhone 快速扫描羽绒服，现场演示生动而充满极客般的探索乐趣。演讲情绪高昂且极具说服力，成功将行业对 AI 替代人类的恐慌与焦虑，转化为重获创作自由与探索全新表达维度的澎湃动力。 ## Key takeaways - [00:04](https://www.youtube.com/watch?v=zuNWlE82ujI&t=4s) 生成式 AI 的核心价值在于消除影视后期流程中的摩擦与繁琐环节，释放创作者的“创意心流”，而非取代人类的艺术构想。 - [02:47](https://www.youtube.com/watch?v=zuNWlE82ujI&t=167s) 11 Dollar Bill 重构了传统的线性后期管线，将剪辑、动态图形、音频与生成式占位素材变为实时并行推进，大幅提前暴露潜在问题并减少补拍成本。 - [04:33](https://www.youtube.com/watch?v=zuNWlE82ujI&t=273s) 工作室将 AI 作为按需索取的灵活“模块化菜单”，不仅重塑了与客户的协作关系，更实现了在剪辑套房内透明、实时的现场共创。 - [07:17](https://www.youtube.com/watch?v=zuNWlE82ujI&t=437s) 借助 Midjourney 与 Runway 等工具快速将构想视觉化，团队有效打破“白纸综合征”，让非专业美术人员与制片人也能精准向视觉特效团队传达抽象概念。 - [09:04](https://www.youtube.com/watch?v=zuNWlE82ujI&t=544s) 利用 Suno 等生成式音频工具，剪辑师能在短时间内为商业广告快速制作情绪高度匹配的定制音乐与音效原型（如 Ocean Spray 案例）。 - [10:51](https://www.youtube.com/watch?v=zuNWlE82ujI&t=651s) 达美乐披萨案例展示了人机协同（Human-in-the-loop）工作流：在获得配音演员授权下利用 ElevenLabs 生成占位音轨用于剪辑迭代，最终依然由真人录制并获得全额酬劳。 - [13:27](https://www.youtube.com/watch?v=zuNWlE82ujI&t=807s) 合成媒体与生成式视频使团队摆脱了图库素材的限制，能够直接实现天气替换（如 Vail Resorts 晴空替换）、镜头构图扩展及器材穿帮智能清除。 - [15:28](https://www.youtube.com/watch?v=zuNWlE82ujI&t=928s) 利用 iPhone 与 Luma 3D 激光雷达扫描，团队在几天内完成了北面羽绒服的 3D 资产建模与帝国大厦实景合成，极速打造爆款社交内容。 - [16:35](https://www.youtube.com/watch?v=zuNWlE82ujI&t=995s) Florastor 益生菌广告实现了端到端的 AI 生产闭环，将真人实拍演员与 Adobe Firefly、Runway 生成的科幻微观人体世界无缝结合。 - [18:08](https://www.youtube.com/watch?v=zuNWlE82ujI&t=1088s) AI 时代催生了具备跨学科能力的“创意通才”（Creative Generalist），极少数掌握 AI 工作流的通才组成的小型特战队即可爆发惊人的创意与执行生产力。 ## Detailed walkthrough ### [00:04](https://www.youtube.com/watch?v=zuNWlE82ujI&t=4s) 重新审视工作流与寻找“创意心流”（Creative Flow）作为后期制作工作室 11 Dollar Bill 的首席创意官兼联合创始人，Christian Robins 在影视后期领域深耕了超过 25 年，并曾斩获多座艾美奖。11 Dollar Bill 业务广泛，在科罗拉多州博尔德、芝加哥以及北卡罗来纳州罗利-达勒姆均设有办公室，曾为达美乐（Domino's）、阿迪达斯（Adidas）、大力水手炸鸡（Popeyes）、JD Sports、Netflix、曼联足球俱乐部（Manchester United）等知名品牌制作商业广告，制作过纪录片，甚至为拉斯维加斯的球形场馆 MSG Sphere 打造过沉浸式内容。Robins 自嘲自己有一种在技术成熟前就“过早采用”的坏习惯，但他的核心职责正是穿透技术炒作的迷雾，去粗取精，将真正行之有效的工具整合到创意工作流中。 Robins 强调，这绝不是一场纯技术的死板宣讲，而是关于如何“炸毁旧工作流”，敢于重塑流程，并最终找到他所谓的“创意心流”（Creative Flow）。在过去几年中，他们已经在真实项目中深度应用生成式 AI。面对许多创意人员对生成式 AI 的抵触与恐慌——担心自己的艺术身份和表达方式被替代——Robins 给出了明确的反驳：如果运用得当，AI 绝不会取代创意构想，反而会极大地放大创意。"我们绝不让生成式 AI 定义我们，而是将其视为回到自由流畅创意状态的途径。" 它的核心价值是消除日常工作中耗时耗力的繁琐环节，保护工艺本身，让创作者将宝贵的时间投入到真正的实验与创新中。 - **11 Dollar Bill 分支机构**：科罗拉多州博尔德（Boulder）、芝加哥（Chicago）、北卡罗来纳州罗利-达勒姆（Raleigh-Durham）。 - **代表客户群体**：Domino's、Adidas、Popeyes、JD Sports、Netflix、Manchester United 及 Sphere 场馆定制项目。 - **核心理念**：AI 消除流程摩擦（Remove friction），保持创作势头（Momentum），释放艺术家的心流体验。 --- ### [02:47](https://www.youtube.com/watch?v=zuNWlE82ujI&t=167s) 线性旧管线与并行新系统的交锋传统的影视后期制作遵循严格的线性流向：实拍结束后，素材进入剪辑，随后依次流转至动态图形设计、音频制作、精修（Finishing），最后打包交付。这种“进-中-出”的单向管道极其僵化，问题往往在流程末期才暴露出来，且几乎没有向上游或下游回溯调整的空间。一旦某个环节需要推倒重来，重新制作构想的成本极为高昂，导致团队陷入无休止的被动等待。为了打破这种束缚，11 Dollar Bill 彻底重构了底层协作系统。新工作流消除了各个环节之间的等待时间，剪辑、设计、动态图形与音频制作全部实现并行推进。团队可以在同一个工作间内随时随地生成临时视频片段（Scratch video）、占位单帧和参考音乐。这一变革让潜在问题大幅提前浮现，决策获得即时视觉佐证，补拍需求显著减少。速度只是该流程的副产品，真正的目的是维系创作心流，并让单兵作战的小团队具备惊人的多维执行力。 ```mermaid flowchart TD accTitle: 传统线性后期管线与全新并行工作流对比 subgraph OldPipeline["传统线性流程（孤岛与迟缓等待）"] O1["实拍拍摄"] --> O2["画面剪辑"] O2 --> O3["图形与特效"] O3 --> O4["音频与配乐"] O4 --> O5["后期精修与交付"] end subgraph NewPipeline["全新并行流程（实时协同与即时反馈）"] N1["前期概念与提示词开发"] subgraph ParallelBlock["实时并行协作区"] N2["剪辑 Editorial"] N3["动态图形 Motion"] N4["生成式音频 / 音乐"] N5["即时草稿视频 / 帧生成"] end N1 --> ParallelBlock ParallelBlock --> N6["实证决策与快速交付"] end ``` --- ### [04:33](https://www.youtube.com/watch?v=zuNWlE82ujI&t=273s) 模块化 AI 菜单与共创式客户关系在 11 Dollar Bill 的日常运转中，AI 并不是强制性的硬性规定，而是一张按需索取的“功能菜单”。有的项目完全不需要 AI 介入，有的只需要单点辅助，而有的则需要五六个模块深度协同。在前期制作阶段，概念构思和脚本开发通过 AI 辅助得以极速推进，极大地压缩了启动周期。这一转变意外地重塑了工作室与客户之间的合作关系。传统的后期制作往往是生硬的甲方乙方买卖关系——客户支付服务费，然后在漫长的等待后验收结果。而当 AI 排除了无效耗时后，双方得以直接坐进剪辑间，面对面实时碰撞想法、共同创作。工作室在工具使用上保持完全透明，将客户教育融入合作流程，甚至在剪辑套房内进行实时现场演示（Live Demo）。涉及声纹或演员肖像生成时，合同中会严格规范授权范围与合规边界。通过提供定制化的解决方案并邀请客户深入上游讨论，团队不仅赢得了高度信任，更在更短时间内交付了更高水准的创意成果。 - **工作流机制**：AI 作为辅助选项（Menu），按项目需求灵活组合，不搞一刀切。 - **客户关系进化**：从单纯的“交付成品”转变为在套房内的“沉浸式现场共创”。 - **透明度与合规**：提前明确数据来源与工具原理，肖像与声音资产严格走授权（Consent & Scope）合同。 --- ### [07:17](https://www.youtube.com/watch?v=zuNWlE82ujI&t=437s) 突破白纸困境：从 Dave & Buster's 到视觉概念即时落地任何创意项目最艰难的一步往往是开始阶段，即令人痛苦的“白纸综合征”（Blank page syndrome）。11 Dollar Bill 突破这一障碍的方法就是：用最快速度把构想具象化。在为大型连锁餐厅酒吧 Dave & Buster's 打造商业广告与视觉升级时，团队利用客户现有的企划方案、现有素材以及尚未实际拍摄的构想，结合 Midjourney 与 Runway 进行快速生成，直接输出了极具质感的故事分镜板。这让导演和各方团队在雇佣任何一名剧组摄制人员之前，就已经对最终视觉基调达成了一致。更重要的是，AI 消除了非专业美术人员表达创意的门槛。过去，如果制作人脑中有一个难以言喻的奇思妙想——比如"让百事可乐像暴雨一样从天而降淋在街道上"，或者"让一个柠檬瞬间变成一只飞鸟"——往往受限于画技或软件技能而无法沟通。现在，团队中的任何成员（包括制片人）都可以直接生成动态预览并递给视觉特效（VFX）团队展示："这就是我想要的效果。" 这种跨职能沟通壁垒的瓦解，彻底解放了团队的想象力。 - **实战案例（Dave & Buster's）**：借助 Runway 与 Midjourney，在摄制组建组前完成高精故事板与视觉定调。 - **沟通门槛消除**：无论职务是否为专职美术，均可直接生成动态样本传达抽象意图。 --- ### [09:04](https://www.youtube.com/watch?v=zuNWlE82ujI&t=544s) 生成式音频的实战威力：从西海岸 Rap 到 Ocean Spray 广告在视觉工具之外，Robins 对生成式音乐和 AI 配音技术情有独钟，称其为一个充满乐趣的“兔子洞”。作为一名 90 年代末期西海岸匪帮说唱（West Coast Gangster Rap）的狂热爱好者，Robins 在演讲前特意用 Suno 的新版本快速制作了一段地道风格的说唱伴奏与曲目，向现场展示了生成式音频极高的律动完成度。而在真实的商业实战中，这种即时音乐生成能力展现了惊人的商业价值。在为知名果汁品牌优鲜沛（Ocean Spray）制作假日社交媒体短视频时，客户突发奇想，希望有一段类似管弦交响乐团加唱诗班合唱（Orchestra choral）风格的短音乐来唱出品牌标语。在传统的后期工作流中，这种临时需求需要寻找版权曲库或耗费数天联系配乐师定制，而剪辑师利用 AI 工具在极短时间内便在套房中敲定并生成了完全匹配情绪的定制音频。这种快速迭代能力让传统后期制作瞬间拥有了前所未有的敏捷度。 - **测试工具**：Suno 音乐生成平台。 - **Ocean Spray 商业案例**：为社交媒体定制短视频，现场直接生成管弦乐/合唱风格的品牌 Tagline 音频，实现即时音频原型制作与快速迭代。 ### [10:51](https://www.youtube.com/watch?v=zuNWlE82ujI&t=651s) ElevenLabs 与达美乐：用 AI 占位配音实现人机协同在商业广告的高频制作中，声音与画面的精准配合至关重要。Christian 分享了他们与 ElevenLabs 的紧密合作，并以知名客户达美乐披萨（Domino's Pizza）为例展示了实际落地场景。达美乐每天都有大量的社交媒体和广播广告上线，其所有广告片都由一位名叫 Casey 的专属配音演员完成。在传统流程中，当剪辑师把画面排好后，往往会发现现有的文案在节奏或语调上不够契合，需要紧急重写。然而，剪辑团队不可能随时要求 Casey 守在麦克风前录制临时的草稿配音（Scratch VO）。为了打破这一时间瓶颈，客户在获得配音演员 Casey 本人完全知情与授权同意的前提下，利用 ElevenLabs 训练了专属的声音模型（团队戏称为“Robo-Casey”）。剪辑师在修改文案后，直接使用 AI 生成占位配音来匹配画面、调整节奏；待脚本和时长完全敲定后，再把干净的最终脚本与节奏参考发送给 Casey，由她录制正式的成片配音。Christian 强调，AI 生成的声音绝不会直接面向消费者投放，它纯粹是内部剪辑辅助工具，“Casey 拿到的报酬一分不少，这让她的工作轻松了十倍，也让我们团队的日子好过得多，为什么不用呢？” 现场播放了 Robo-Casey 制作的 9.99 美元披萨优惠草稿，随后播放了 Casey 录制的正式版本（Christian 还打趣说达美乐现在又推出了 6.99 美元的新特价）。他指出，这就是人机协同（Human-in-the-loop）的典范——在设立安全边界的同时解放创意团队的时间，将原本需要一整天等待返工的死板流程，压缩到短短几小时内自由尝试多种不同文案和玩法，找回做广告创作本身的纯粹乐趣。 ```mermaid flowchart LR accTitle: 达美乐人机协同配音工作流 subgraph S1["剪辑与文案迭代"] A["剪辑师排布画面"] --> B["调用 Robo-Casey 生成占位音轨"] B --> C{"节奏与语调是否匹配？"} C -- "不合适" --> D["快速改写文案并重新生成"] D --> B end subgraph S2["真人交付与终审"] C -- "完美契合" --> E["发送定稿脚本与参考音轨给真人配音"] E --> F["真人 Casey 录制最终音轨并获全额酬劳"] F --> G["最终商业成片出街"] end ``` - 核心原则：在获得演员明确授权的前提下训练 AI 声音模型，仅用于内部剪辑（Scratch VO），不替代最终人类配音。 - 业务成效：极大减少了演员反复试录琐碎草稿的无谓消耗，赋予剪辑师即时测试不同创意脚本的自由度。 --- ### [13:27](https://www.youtube.com/watch?v=zuNWlE82ujI&t=807s) 生成式视频与画面处理：从天气替换到智能去穿帮谈到合成媒体（Synthetic Media）与生成式视频，Christian 难掩兴奋，直言这是他目前最痴迷的领域。在传统后期中，创意人员往往要在图库网站上耗费数小时寻找合适的素材片段，这无异于巨大的时间黑洞。而生成式工具的出现彻底改变了这一现状——“你为什么不直接动手做出来？想做成什么样就做成什么样。” 团队可以借助 AI 对实拍素材进行延伸、修改与操控，例如修正宠物在片场不听指令走位、或者演员没有看镜头的遗憾。 Christian 以大客户韦尔度假村（Vail Resorts）为例，该品牌曾斥资数十万美元进行实地外景拍摄，渴望捕捉到万里无云的绝美晴空（Bluebird sunny day），但无奈天公不作美，现场只拍到了阴沉灰暗的阴天。团队利用生成式 AI 技术将阴沉的画面直接转换为明媚的艳阳天，不仅视觉质感极其震撼，还为品牌方挽回了昂贵的重拍预算。此外，AI 工具在日常“低垂的果实”（Low-hanging fruit）任务中同样展现了惊人效率： - **画面扩展与器材擦除**：当团队对 6K 或 8K 高清素材进行缩放构图调整时，画面边缘往往会暴露出 C 型支架（C-stands）、灯光设备与杂乱线缆。过去这种穿帮镜头会一直挂在团队心头，直到下周进入专门的清理流程；如今剪辑师在几秒内就能一键将其清理干净，彻底消除心理负担。 - **无绿幕场景替换**：无需预先搭建昂贵的绿幕背景，直接通过 AI 算法精准抠像并替换环境背景，将过去繁重复杂的后期工种变成了日常轻量化操作。 --- ### [15:28](https://www.youtube.com/watch?v=zuNWlE82ujI&t=928s) Luma 3D 扫描与敏捷社交内容：帝国大厦羽绒服案例在快节奏的社交媒体营销中，速度往往决定了一切。Christian 分享了他们为零售品牌 JD Sports 制作的一则爆款案例：品牌方需要为北面（The North Face）羽绒服紧急制作一支极具冲击力的社交宣传短片。团队并没有采用耗时数周的传统 3D 建模流程，而是采取了一种极其敏捷且充满趣味的做法：在寒冬的制片办公室里，团队找来一件北面羽绒服，往里面塞满了亚马逊快递纸箱让它饱满定型，看起来就像是坐在帝国大厦顶端一样。随后，团队拿出 iPhone，使用 Luma 3D 的激光雷达（LiDAR）扫描功能绕着衣服快速扫了一圈，直接生成了高精度的 3D 资产模型，紧接着将模型跟踪合成到了纽约帝国大厦的实景镜头上。 - **制作周期对比**：传统三维建模与追踪合成需要耗费数周时间，而借助 Luma 3D 工具，整个创意从构思到完成仅用了几天时间。 - **传播效果**：该短片上线后在社交媒体上斩获了数百万次播放。 - Christian 特别提到，许多人目前依然低估了 Luma 的技术实力，并推荐现场观众去参加 Luma AI 创始人 Amit Jain 次日举行的炉边对话。 --- ### [16:35](https://www.youtube.com/watch?v=zuNWlE82ujI&t=995s) 整合全流程落地：Florastor 益生菌的微观世界广告当团队将上述小型工具串联起来时，便能迈向大型复合级 AI 制作。保健品牌 Florastor 希望成为益生菌行业内首个将生成式 AI 视觉应用于正式商业广告的品牌。11 Dollar Bill 与该品牌的代理公司深度协作，利用 Adobe Firefly 和 Runway 进行了前期的概念设计、视觉探索与分镜预演。在实拍阶段，团队拍摄了绿幕真人演员素材，而在环境搭建上，则完全依托“图生视频”（Image-to-Video）技术，构建出了一个充满未来感与科幻质感的微观人体肠道细菌城市。在这个由 AI 生成的环境中，“益生菌卫士”正在全力清除有害菌、增强有益菌群。Christian 现场播放了这支成片，画面唯美流畅，且整支广告中除了真实演员是实拍之外，所有宏大的微观背景全部由 AI 赋能生成。这一案例印证了从前期构思、中期生成到后期合成的完整 AI 生产闭环。 --- ### [18:08](https://www.youtube.com/watch?v=zuNWlE82ujI&t=1088s) 定制 AI 基础设施与“创意通才”小团队的崛起在总结发言中，Christian 阐述了其咨询机构 Agent11 正在探索的终极方向：针对传统后期软件无法解决的独特创意痛点，为客户量身定制专属的 AI 工作流与创意 Agent。这种新型的基础设施并非由纯技术极客开发，而是由深谙影视语言与后期运动规律的资深创作者打造，旨在不牺牲人类品质的前提下，将创作者的情感与直觉深度注入到流程之中。 ```mermaid flowchart TD accTitle: 创意通才的技能与赋能模式 subgraph G["创意通才 (Creative Generalist) 架构"] Core["核心专业强项 (剪辑 / 导演 / 视觉设计)"] Ancillary["广泛的辅助技能 (调色 / 声音 / 3D / 文案)"] Core --- Ancillary end AI["生成式 AI 工具链 (消除摩擦与机械性痛点)"] --> G G --> SuperTeam["敏捷无敌的小型创意团队 (Unstoppable Small Teams)"] ``` Christian 借用了好友、Signal & Cipher 联合创始人 Brent Peterson 提出的概念——**“创意通才”（Creative Generalist）**： 1. **多学科融合**：未来的顶尖创作者是在某一领域具备核心专业深度、同时在其他相关领域（如调色、合成、3D、音频）拥有广泛辅助技能的复合型人才。 2. **消除摩擦与说“不”**：AI 工具消除了日常流程中那些令人抓狂的琐碎障碍，让创作者在面对前所未有的创意构想时“永远不必说不，而是直接做出来”。 3. **小团队的无限威力**：当几个充满好奇心、熟练掌握各种 AI 工具的创意通才组合在一起时，就能构成战斗力极强、势不可挡的小型创意特战队。演讲最后，现场播放了一段由 AI 生成的女声致辞：“嗨，Ai4 的朋友们，感谢你们今天听我唠叨生成式 AI，这真的很酷！” 在全场热烈的掌声中，主持人高度评价这是他“整天最喜欢的精彩演讲”，Christian 亦向全场致谢并鼓励大家在拉斯维加斯的大会中持续探索前沿变革。 ## Notable quotes > [00:04](https://www.youtube.com/watch?v=zuNWlE82ujI&t=4s) 我们绝不让生成式 AI 定义我们，而是将其视为回到自由流畅创意状态的途径。 — Christian Robins > [00:04](https://www.youtube.com/watch?v=zuNWlE82ujI&t=4s) 炸毁旧工作流 — Christian Robins > [07:17](https://www.youtube.com/watch?v=zuNWlE82ujI&t=437s) 让百事可乐像暴雨一样从天而降淋在街道上 — Christian Robins > [07:17](https://www.youtube.com/watch?v=zuNWlE82ujI&t=437s) 让一个柠檬瞬间变成一只飞鸟 — Christian Robins > [07:17](https://www.youtube.com/watch?v=zuNWlE82ujI&t=437s) 这就是我想要的效果。 — Christian Robins > [10:51](https://www.youtube.com/watch?v=zuNWlE82ujI&t=651s) Casey 拿到的报酬一分不少，这让她的工作轻松了十倍，也让我们团队的日子好过得多，为什么不用呢？ — Christian Robins > [13:27](https://www.youtube.com/watch?v=zuNWlE82ujI&t=807s) 你为什么不直接动手做出来？想做成什么样就做成什么样。 — Christian Robins > [18:08](https://www.youtube.com/watch?v=zuNWlE82ujI&t=1088s) 永远不必说不，而是直接做出来 — Christian Robins > [18:08](https://www.youtube.com/watch?v=zuNWlE82ujI&t=1088s) 嗨，Ai4 的朋友们，感谢你们今天听我唠叨生成式 AI，这真的很酷！ — AI 生成声音 ## Mentioned Christian Robins, 11 Dollar Bill, Agent11, Ai4, Domino's (达美乐), Adidas (阿迪达斯), Popeyes (大力水手炸鸡), JD Sports, Netflix, Manchester United (曼联), MSG Sphere, Dave & Buster's, Midjourney, Runway, Suno, Ocean Spray (优鲜沛), ElevenLabs, Casey, Robo-Casey, Vail Resorts (韦尔度假村), The North Face (北面), Luma 3D (Luma AI), Amit Jain, Florastor, Adobe Firefly, Brent Peterson, Signal & Cipher ## Video description Christian Robins, CCO/Co-Founder, 11 Dollar Bill and Agent11 Abstracts Generative AI is transforming advertising, empowering AI-Enhanced Creatives to work faster, iterate seamlessly, and push the boundaries of storytelling. In this session, Christian Robins, multiple Emmy Award-winning creative and co-founder of post-production studio 11 Dollar Bill and GenAI consultancy Agent11, shares his practical, workflow-driven approach to integrating generative AI into real-world creative processes. Christian will break down small, medium, and big steps for AI adoption - demonstrating how tools like AI-powered ideation, generative video, and synthetic media are revolutionizing creative workflows. Learn how AI can enhance collaboration, remove bottlenecks, and accelerate content creation pragmatically - not replacing human creativity, but amplifying its potential. This session explores how AI-Enhanced Creatives - multidisciplinary experts who merge AI tools with traditional creative skills - are leading the next wave of advertising innovation. Walk away with actionable strategies to harness generative AI, refine your creative process, and unlock new possibilities in modern advertising. [Source](https://www.youtube.com/watch?v=zuNWlE82ujI)

Source quotes

[0:04] Christian Robins: 我们绝不让生成式 AI 定义我们，而是将其视为回到自由流畅创意状态的途径。 [0:04] Christian Robins: 炸毁旧工作流 [7:17] Christian Robins: 让百事可乐像暴雨一样从天而降淋在街道上 [7:17] Christian Robins: 让一个柠檬瞬间变成一只飞鸟 [7:17] Christian Robins: 这就是我想要的效果。 [10:51] Christian Robins: Casey 拿到的报酬一分不少，这让她的工作轻松了十倍，也让我们团队的日子好过得多，为什么不用呢？ [13:27] Christian Robins: 你为什么不直接动手做出来？想做成什么样就做成什么样。 [18:08] Christian Robins: 永远不必说不，而是直接做出来 [18:08] AI 生成声音: 嗨，Ai4 的朋友们，感谢你们今天听我唠叨生成式 AI，这真的很酷！

Source quotes

[0:04] Christian Robins: bad habit of early adopting tech before it's ready [0:04] Christian Robins: Keep momentum, protect the craft, and then let AI remove the friction so we can all kind of stay in flow. [4:33] Christian Robins: a group of creative people in a room working together. [9:04] Christian Robins: Casey gets paid the same. Makes her life a hell of a lot easier. It makes our life a hell of a lot easier. Why wouldn't you be using these tools? [17:10] Commercial Voiceover: It's another superb day in Gut City thanks to the flora restorers working hard to flush out the bad and boost your own good bacteria. Thanks to them, your gut is more balanced and resilient with Florastor probiotics. It's all systems go. [17:10] Christian Robins: The only thing real on that spot are the people. Everything else is for fun. [18:08] Christian Robins: Anybody who has a strength and has all these ancillary strengths now can take these tools, eliminate the nightmares out of their day, and then surround themselves with people who are also interested in all different types of content and tools... and then you have these small teams that are totally unstoppable now. [19:33] Synthetic Avatar: Hey there, Ai4. Thanks for listening to me ramble today about generative AI. This is pretty neat. [19:33] Christian Robins: And she's right. This is pretty neat [19:33] Session Moderator: Everybody give it up for Christian Robins. That was easily my favorite presentation all day. I must say that was awesome.

Original transcript

[0:04] My name is Christian Robbins. I'm chief [0:05] creative officer and co-founder of $11 [0:08] bill. And what Bill is, it's a [0:11] post-production studio. So we make TV [0:13] commercials [0:15] um for Dominoes, Adidas, Popeye's, JD [0:19] Sports, Netflix, Manchester United. [0:21] We've done spots for everybody over the [0:23] last couple of decades. We've got [0:25] offices in Boulder, Colorado, Chicago, [0:29] and Raleigh, uh, Durham, North Carolina. [0:31] Um, we've made documentaries. Um, we [0:34] made some that's on the sphere right now [0:36] if you look down the street. So, [0:37] essentially, we kind of never say no to [0:39] anything. We make whatever we want. Um, [0:43] and like I said, I've been doing post [0:45] probably a little over 25 years now. uh [0:48] won some shiny awards and I've got a bad [0:51] habit of early adopting tech before it's [0:53] ready. Um but that's my job. It's to [0:55] curate what works. We sift through the [0:57] BS and implement it into our creative [1:01] workflow. And you know that's what this [1:04] talk is about. Um build it's it's [1:06] blowing up your old workflow, taking a [1:08] hard look at it, reinventing it, and not [1:10] being afraid to reinvent it. All in the [1:12] end to find something I call creative [1:15] flow. And I'm sure you're familiar with [1:16] the creative flow state, just that, you [1:19] know, nice place to be creatively. And [1:21] I'll get I'll get there in just a [1:22] minute. But this is not a technical talk [1:24] by any means. Um, so it's about me [1:27] showing you actual pragmatic real world [1:30] application of generative AI in uh [1:33] post-production. So you know what is [1:35] creative flow? Um, like I said, we've [1:38] been using gen and our genai and our uh [1:42] real workflows for a few years now. Um, [1:45] and you know, a lot of what I'll show [1:46] through you is show you today is through [1:48] a post-production lens, but there's [1:50] something to take away for everybody. [1:52] Um, and I think like what this really [1:53] comes down to is a lot of creatives feel [1:55] threatened by generative AI. And like I [1:58] get it, you know, our work is tied to [2:00] our identity creatively. You know, the [2:02] way we execute is tied to the way, you [2:04] know, our art. So here's my take. I [2:08] think that generative AI is just [2:09] completely different. [2:11] um it doesn't replace creative vision [2:13] and in fact you know it amplifies it if [2:16] done right. So again we don't let [2:18] generative AI define us by any means. [2:21] You know we see it as a way to get [2:22] ourselves back to this fluid state of [2:25] creative freedom. So you make the time [2:27] consuming parts of the day go away so [2:30] you can spend more time being creative [2:32] and experimenting. Um that's the [2:35] creative flow state. So that's the goal, [2:38] right? keep momentum um protect the [2:40] craft and then let AI remove the [2:43] friction so we can all kind of stay in [2:44] flow. So you know here's a quick look at [2:47] what the old pipeline looks like and you [2:51] know why this breaks flow. So [2:54] post-production, like I said, is [2:55] everything that happens after the shoot, [2:56] right? You know, we have a production, [2:58] we film people, we bring in editing, uh [3:01] we bring in footage to edit with, we do [3:03] graphics, and then we do audio, then we [3:05] do finishing, and then we do delivery, [3:07] right? It's an in mid out very linear [3:11] process. [3:12] Um, and the issues are simple, right? [3:15] Like this moves really slowly. There's [3:16] not a lot of room for it to go upstream, [3:18] downstream. Um, you know, problems [3:21] surface late. uh changes are expensive [3:25] and then you know re-executing ideas, [3:27] concepts when it's not just right and [3:29] then you hopefully have it final and [3:31] then you deliver it. So that's the trap [3:35] of the old workflow. So we rebuilt the [3:38] system at $11 bill and I promise I don't [3:42] have a whole ton of charts or anything. [3:44] It'll be a lot more fun but um you kind [3:46] of see what happens, right? Everything [3:49] stops waiting. um editorial design, [3:53] motion graphics, audio all happen at the [3:56] same time. Um scratch videos, frames, um [4:00] temp music, we can do all this in the [4:03] room whenever we decide to now. So the [4:06] problems pop up sooner. Um decisions get [4:10] evidence, right? And pickups drop. So [4:14] speed is the byproduct of this and the [4:16] goal is keeping that creative flow and [4:19] all these things. now can be done by one [4:22] person which makes small teams crazy [4:26] crazy advanced in what they can do and [4:29] I'll show you exactly where that plugs [4:30] into our pipeline dayto day. Um, this is [4:33] kind of our menu, right? It's not a [4:35] mandate, but these are the things that [4:38] every part of our pipeline has a [4:40] generative AI assist if we want it. Um, [4:43] some jobs don't need any, some need one [4:46] piece, some need five or six, but [4:49] you know, the point is to keep the flow [4:51] going. So, we pick the combo of these [4:55] generative AI helpers uh to, you know, [4:58] move the work and just skip the ones [5:00] that we don't need to worry about. [5:03] um you know pre-pro is you know [5:05] pre-production it has turned into [5:07] concept and script development obviously [5:09] it will work with people but ideulating [5:11] a lot of that makes it move much much [5:14] quicker than it has in the past um and [5:17] in the end like here's a fun byproduct [5:18] of the whole thing you just overall [5:20] elevate the entire creative experience [5:23] because now when you remove all the time [5:25] sucks the creative experience gets [5:27] better for everybody and our teams stay [5:30] in flow and then it becomes less less of [5:32] a vendor client relationship and more of [5:35] a group of creative people in a room [5:38] working together. And our clients [5:41] generally aren't used to that. You know, [5:43] we offer a service, they pay us to do [5:45] the service for them. If we can all get [5:47] in the room, be creative together, [5:49] that's a really mindopening thing for a [5:51] lot of our client base and they feel [5:53] like they're part of the process, not [5:54] just stuck waiting to see what happens, [5:57] right? And you know in the end the [5:59] results trust you know and momentum and [6:02] better work in less time. So we stay [6:05] transparent about the tools and you know [6:06] this is the part of it that you know [6:08] before I go any further in this [6:10] conversation we make education part of [6:12] the process because you know obviously [6:14] we've spent a lot of time trying to [6:15] figure out these tools the clients may [6:17] may or may not know how to do that. So [6:20] we get buy in, you know, we do [6:23] we use these tools to uh with our [6:26] clients, not to the clients. Um we open [6:29] the door early. We show them what tools [6:30] are, how they work, where it's being [6:32] sourced from. Um keep guard rails very [6:35] simple. Uh you know, and we demo with [6:38] them live in the suite in a different [6:40] kind of capacity than we ever had [6:41] before. um you know we'll use a voice or [6:44] likeness and you know consent and scope [6:47] are in the contract for that and talent. [6:49] Um and some uses are just editorial [6:52] content only but you know in the end we [6:54] never you know we we don't hand them [6:56] like a one pager. We say hey this is a [6:59] custom solution I think we can work with [7:00] you on and it keeps us all being [7:02] creative. We want to see how it looks. [7:03] How comfortable are with are you with [7:05] it? How comfortable is your client with [7:06] it? And then that opens up a whole line [7:09] of conversation that clients aren't [7:11] generally used to when you're going [7:13] upstream. Um, so this is how we start. [7:17] We start with small steps with our [7:18] clients. So ideation and you know the [7:20] hardest part about any project is just [7:23] starting right like blank page syndrome [7:26] really sucks and we break it by making [7:29] something fast. So these, you know, this [7:32] is for Dave and Busters. Has anybody [7:33] been to a Dave and Busters before? Yes. [7:35] I've got two daughters and every time I [7:38] go it's 250 bucks minimum. But so we [7:40] know how how it works, how it looks in [7:42] there. And they were redesigning their [7:45] uh their source and making some [7:46] commercials out of it. So we took their [7:49] treatments, we took um existing imagery, [7:51] imagery that hadn't even been made yet [7:54] and put it all through runway and [7:56] midjourney and came up with some really [7:58] legit looking storyboards for everybody [8:01] to see. And you identify a vision right [8:04] off the bat. you get the director right [8:05] in the mix right off the bat. Everybody [8:08] knows what's happening before you've [8:09] hired a single crew person to execute [8:12] this concept. So like and that's [8:14] something I think you've everybody's [8:15] kind of seen before. But um you know the [8:19] one thing I want to talk about is [8:20] ideiation [8:22] beyond that. So here's the thing. [8:26] Anybody who has a creative idea, it [8:28] doesn't matter if they're creative or [8:30] not now has access to show people what [8:32] the hell you're talking about, right? [8:34] So, you know, you ever have an idea [8:36] that's totally impossible to explain and [8:38] you just wish you had the skill set to [8:40] show somebody what you mean? Well, you [8:41] know, like here, no, I mean like make [8:43] the Pepsi rain from the sky into the [8:44] street or, you know, let's have this [8:46] lemon turn into a bird, right? So, I can [8:48] just make that or anybody, the producer [8:50] on our team, anybody can make these [8:52] things and show it to the VF VFX team [8:54] and say, "This is what I want." [8:58] So, this barrier for entry is nearly [9:00] eliminated if you know how to work the [9:02] tools. [9:04] And uh you know, here's some fun stuff. [9:06] I really dig um that generative music [9:09] and voice over. I think it's such a cool [9:11] rabbit hole. You can go down your free [9:13] time. Um so if you know me um you know [9:17] that I'm a late 90s gangster rap [9:20] afficionado, West Coast. Um so I made a [9:23] song at Sunno right before this [9:24] presentation. [9:25] >> Yeah. In that style. So yeah, [9:29] it's got a beat to it. [9:32] But this is done with their new version [9:35] lights. [9:37] I confess [9:39] all [9:41] the [9:44] discredit. [9:50] >> So that's fun. But how do we actually [9:53] use this? Right. Um so here's a [9:56] practical application of it. We, you [9:59] know, we did some, uh, holiday spots for [10:01] Ocean Spray. Um, client had a wild ass [10:04] idea to try, what if we just made a [10:05] track, like just made a little stab, and [10:07] this is just a quick social media [10:09] content video. Um, and we had like a [10:12] orchestra coral type of thing, sing the [10:16] tagline. And I was like, okay, let's try [10:17] it out. So, we did. And here's what it [10:19] looks like. [10:23] [Music] [10:28] [Applause] [10:31] So, that's pretty cool, right? Like, the [10:34] fact that you could just knock out a [10:35] custom song real quick for your client [10:37] is pretty groundbreaking, especially [10:39] when you're in a traditional [10:41] post-production editorial situation. So, [10:44] the editor can just make this real quick [10:46] now. Um, and we can rapid, you know, [10:49] rapidly iterate it, right? So, the fact [10:51] that we can do a lot of this is really [10:53] cool. And if you've been around and you [10:54] meet the uh 11 Labs guys, um we love [10:58] them so much. They're doing some really [10:59] fun things. And one of our one of our [11:01] clients uh is Domino's and we do a ton [11:05] of Domino's pizza work, social content, [11:07] broadcast. We're knocking out spots [11:10] every every day for them. So they have [11:14] one VO talent for all of their spots, [11:17] social content, otherwise. Her name's [11:19] Casey. She's awesome. So when we get a [11:22] project, we have a basic set of [11:23] direction, some footage, a script, and [11:26] when you start to lay low as its [11:28] picture, sometimes the script doesn't [11:29] work out all the way. So then you got to [11:31] do quick, you know, quick rewrites, you [11:34] know, maybe change the tone and the [11:35] pace. So the VO read, we can't wait for [11:36] Casey to be sitting around, you know, to [11:39] give us another scratch VO record. So [11:41] our client trained an 11 Labs model with [11:44] the talent's consent. Um, and we use it [11:47] for scratch for everything. So, we apply [11:49] these to picture um block pacing and [11:52] then send a clean script with an idea of [11:54] tone and pacing to Casey and Casey reads [11:57] it. Um, and this never sees, you know, [11:59] the consumer at all. These are just for [12:01] editorial sake. So, Casey gets paid the [12:03] same. Makes her life a hell of a lot [12:05] easier. It makes our life a hell of a [12:06] lot easier. Why wouldn't you be using [12:08] these tools? So, here's a rough cut with [12:11] we technically call her robo Casey just [12:13] so you can kind of hear that how it [12:14] works. What about our customizers? [12:16] Turning our pizza into your masterpiece. [12:19] Good news. Right now, get any pizza on [12:21] any crust with any toppings for $9.99 [12:23] each online. [12:25] >> So, that's not bad for a scratch video, [12:27] Reed. That's pretty damn good, actually. [12:29] And here's the final. [12:31] >> Those of you who turn our pizza into [12:33] your masterpiece, [12:34] >> this is a perfect pizza. [12:35] >> Good news. Right now, get a pizza on any [12:37] crust with any toppings for $9.99 each [12:40] online. [12:41] >> And now it's actually $6.99. they have a [12:42] new thing, a new special out. But um [12:45] regardless, that's the those are the [12:47] guardrails, right? Like we keep the [12:48] human in the loop and we make [12:50] everybody's life easier and everybody [12:52] continues to make money. So if there's a [12:55] way we can still bottle up this workflow [12:58] and make it easier for us to be [13:00] creative, right? What does this do? This [13:02] saves time for us to try out new [13:03] scripts. Hey, I've got a cool idea. [13:05] Let's try it out. Now we've have like [13:07] hours instead of this compressed time [13:09] frame of a day to knock out even more [13:12] content and try things out and just see [13:15] what happens. Like this is a fun [13:16] industry we're in. I think people forget [13:19] that. Um making commercials is pretty [13:21] cool. Like you get to work with some [13:23] really neat people. Let's remember how [13:24] to make it fun again. And this is how we [13:26] do it. [13:27] Um and then my favorite part about, you [13:31] know, uh AI these days is that synthetic [13:33] media, the generative video. I love this [13:35] stuff. so much. Um, I could go on and on [13:38] about this for hours, but there's really [13:40] neat applications of this tech. Like, [13:43] one of the dumbest things is having to [13:45] waste time on a stock website looking [13:47] for something to use. It's a massive [13:50] time suck. So, why don't you just make [13:52] it? You make what you want, right? Super [13:56] easy. So, let's say you take some [13:58] existing IP of yours, shoot like uh, you [14:00] know, the dog kept walking off the [14:02] stage. we told him to sit or we we [14:04] really wish that guy would look in the [14:06] camera. We told him to look into the [14:07] camera. So these tools are out there for [14:10] everybody to use, you know, and we [14:12] extend this footage, right? And and [14:14] manipulate it. This is a great example. [14:16] Bill Resorts, they're a big client of [14:18] ours. They spend hundreds of thousands [14:19] of dollars on production hoping that it [14:22] looks like a bluebird sunny day. And [14:25] that doesn't always happen. Mother [14:26] nature kind of doesn't like to do this. [14:28] So, what we did is showed them what it [14:31] would look like if we made it a sunny [14:33] day. And when you look at what we can do [14:36] now with these tools, this is pretty [14:38] incredible stuff. So, where would you [14:41] rather spend your money at, right? The [14:43] sunny one or the overcast gray looking [14:45] one. [14:47] And then, you know, some lowhanging [14:49] fruit. This was a big time saver for us. [14:51] You know, clean up. Like sometimes you [14:53] want, you know, you get footage that's [14:54] like 6 8K. You can kind of reorganize it [14:57] and play around with the looks. So, [14:59] let's say we wanted to zoom out and get [15:00] a better shot. Well, now you expose [15:02] things like seastands and lighting [15:04] equipment and cables. Cool. Just get rid [15:06] of it. Like that's something that would [15:08] be in the back of everybody's head until [15:10] the following week when you do all the [15:12] cleanup. Just have the editor do it [15:14] right now and then we all just move on [15:15] in our day and that's not rotting at [15:17] your brain. Um, simple stuff like scene [15:20] replacement, right? None of this was [15:21] green screen. We just changed out [15:23] background. like that's not a huge job [15:25] for anybody anymore. [15:28] And this is really cool, too. Um, uh, so [15:31] we do a lot of work with JD Sports and [15:33] they want to do a quick turnaround [15:35] social media piece. So they sent us over [15:38] a Northface jacket that they wanted to [15:40] sell. Um, we stuffed it full of Amazon [15:43] boxes to look like it was sitting on the [15:45] Empire State Building. And then we used [15:48] Luma 3D to do a LAR iPhone scan and [15:52] just, you know, ran around our [15:53] producers's office during the middle of [15:54] the winter and got ourselves a model out [15:56] of it and then took it and tracked it on [16:00] to the Empire State Building and it got [16:02] millions of hits. It was crazy just how [16:05] simple this stuff is. And you know, [16:07] Luma's people are sleeping on Luma right [16:09] now. Lumin is doing really great stuff [16:10] and uh tomorrow if you get a chance to [16:12] see Amit Jane um he's got a fireside. [16:15] He'd be he'd be a great guy to listen [16:17] to. So this is what I'm saying. We can [16:19] create whatever we want now, right? This [16:22] is blowing up your workflow and this is [16:24] achieving pure creative flow. You can [16:27] just come up with anything you want. [16:28] This process would have took weeks to [16:30] do. Took us a like couple of days for [16:33] fun. So then you start to go a little [16:35] bit bigger steps, right? like let's see [16:37] what happens when we start putting all [16:38] these pieces together. Um you know we [16:41] ideulate with our client now we have [16:43] this you know Flora store is one of our [16:45] clients um they really wanted to get [16:47] into the generative AI space with [16:49] visuals and be kind of first to market [16:51] with their advertising. Nobody's really [16:53] been doing that yet in that very [16:55] specific niche. So they were like let's [16:57] do you know let's work together. So we [16:59] worked with their agency, came up, [17:01] played with uh um Firefly and Runway and [17:04] came up with all these designs and [17:05] concepts, these executions, did a [17:08] traditional green screen shoot knowing [17:10] that we put these people in, you know, [17:12] these environments that we created and [17:15] then, you know, start doing some uh [17:18] image to video stuff out of it, right? [17:20] So now we have a fully fleshed out [17:22] environment that looks amazing and we [17:25] can and has kind of an ownable look and [17:28] you know start to composite our people [17:30] in there. So I'll just show you the spot [17:32] real quick. It's uh it's a shorty but [17:35] >> it's another superb day in in city [17:37] thanks to the flora restoers working [17:39] hard to flush out the bad and boost your [17:41] own good bacteria. Thanks to them your [17:44] gut is more balanced and resilient with [17:46] Florestore probiotics. It's all systems [17:48] go. So, you know, that's pretty cool, [17:51] right? Like the only thing real on that [17:53] spot are the people. Everything else is [17:54] for fun. So, this is the loop, that [17:57] workflow loop I told you. This is it in [17:59] action. You're able to do every part of [18:01] it. Um, and you know, continually refine [18:04] this into a workflow that's unique for [18:06] every project. And I know we got to wrap [18:08] this up soon here. So, um, you know, [18:11] just overall complete steps. This is [18:13] what we're doing now. We're building [18:14] custom AI workflows for unique creative [18:17] problems. So trading you know agents [18:21] with our creative identities. We're [18:22] building workflows for a challenge of [18:24] each project where the traditional tools [18:26] from post-production can't even touch. [18:29] So it's creative infrastructure we're [18:30] building now that is built by people who [18:33] know how to move that without [18:34] compromising the way they work and [18:36] channel all their human qualities into [18:39] the workflow. And this is a really [18:41] important thing to kind of leave on [18:43] here. So this statement is something [18:45] really important for everybody to [18:46] realize. We combine all these tools in a [18:49] way to like reduce friction and create a [18:51] seamless environment where we never have [18:53] to say no. We can just do it. It's [18:56] great. And the more experimentation, you [18:59] know, you become multiddisciplined, [19:01] which turns you into a creative [19:03] generalist. And this is a a term my [19:05] buddy Brent Peterson at uh Signal and [19:07] Cyper came up with and I I stole it. I [19:10] put it in every presentation I do [19:11] because I love that term. So anybody who [19:13] has a strength and has all these [19:15] ancillary strengths now can take these [19:17] tools, eliminate the nightmares out of [19:19] their day and then surround themselves [19:22] with people who are also interested in [19:24] all different types of content and tools [19:26] and things along that way. And then you [19:28] have these small teams that are totally [19:30] unstoppable now. It's it's really neat. [19:33] So again, just to recap, I've shown you [19:36] a lot of stuff, but if you're trying to [19:37] implement this workflow change at your [19:40] company, small, medium, large, and [19:43] that's how you do it, one one workflow [19:45] shift at a time. And [19:48] >> hey there, AI4. Thanks for listening to [19:50] me ramble today about generative AI. [19:52] This is pretty neat [19:53] >> And she's right. This is pretty neat [19:56] So So yeah. So anyway, thank you all for [19:59] listening me ramble. Um, you have a [20:01] great time in Vegas. We're all learning [20:03] so much here. [20:05] >> Everybody give it up for Christian [20:06] Robbins. That was easily my favorite [20:08] presentation all day. I must say that [20:10] was awesome. [20:11] >> Thank you. Thank you everybody.