20 January 2026

233M Views in 3 Days: The David Beckham AI Workflow

2Frameworks

12Insights

Frameworks in this episode

Innovation6 steps

Five-Role AI Video Production Pipeline

Split AI filmmaking into five specialist roles and move approved shots downstream.

Read →

Innovation6 steps

Two-by-Two Consistency Storyboarding

Generate four related frames together, then crop and sequence them into a scene.

Read →

Insights & moments

The myth-busts, hot takes, explainers, and tools worth keeping.

Hot Take· 3

Hot Take22:00

The Future Acting Advantage Is Range, Not Appearance

PJ argues that AI will not eliminate nuanced acting but will let small casts drive many digital characters. Faces and voices can be reskinned, making vocal flexibility, accents, micro-expressions, and performance range more valuable than possessing a particular screen look.

Prompting alone does not reproduce nuanced human performances.
Small teams of actors may drive large digital casts.
Faces, apparent ages, and voices can all be transformed.
Voice actors may be unusually well prepared for this production model.

“We don't get the nuance, we don't get the micro expressions but we are going to have it Avatar style where a small team of…”
PJ Ace · 22:00

“It's not if you look like Timothy Chalamet or Sydney Sweeney, it's just what's your range as an actor.”
PJ Ace · 23:30

#acting#voice acting#digital characters#future of film

Hot Take24:00

AI Will Replace Mass-Market Constraints With Hyper-Niche Stories

PJ predicts that lower production costs will let filmmakers create highly specific stories for small fandoms rather than aiming every project at a broad television audience. He expects professional filmmakers to improve the medium while also producing a large volume of low-quality work.

Early AI content looks weak partly because many professional filmmakers have avoided the tools.
Commercials provide a less culturally sensitive entry point than beloved film properties.
Internet distribution rewards narrow interests rather than six-channel mass appeal.
Niche creators will expand established fictional universes with highly specific stories.
Lower barriers will produce both exceptional work and substantial junk.

“Of course it's slop. Not many professional filmmakers have like used these tools”
PJ Ace · 24:00

“Like it's just going to be this bright bright future and a ton of crap.”
PJ Ace · 25:30

#niche media#storytelling#filmmaking#creator economy

Hot Take39:30

Why Challenger Brands Have an AI Advertising Advantage

PJ argues that challenger brands often outperform incumbents with AI advertising because audiences have no sacred visual tradition to defend. Established brands must meet existing expectations, whereas newcomers can use unusual AI creative to surprise people and establish initial awareness.

Incumbents inherit strong audience expectations and brand traditions.
AI reinterpretations of beloved properties can provoke backlash even when technically competent.
Challenger brands have fewer preconceived creative constraints.
Novel AI work can help an unknown brand become memorable.
Large incumbents must invest more heavily in realism, craft, and brand safety.

“challenger brands like Kelshi are by far the ones that do the the the best with AI because no one has preconceived expectations on what…”
PJ Ace · 40:00

“if there's no frame of reference, you can kind of use this as a new way to build awareness, build that initial understanding.”
Host · 40:30

#challenger brands#brand strategy#ai ads#brand awareness

Explainer· 2

Explainer30:00

Why Film Production Will Go Hybrid Before It Goes Fully Digital

PJ expects near-term productions to retain real actors while using AI for the wide shots, battle scenes, and spectacle that drive budgets upward. He compares this model to Avatar and virtual-production workflows, where real performances remain central even when the surrounding movie is digital.

Actors can perform on LED stages while AI supplies expensive spectacle.
Wide shots and battle sequences offer especially large savings.
Hybrid production can retain union performers and human emotional nuance.
Fully digital films may become acceptable when real actors still drive the performances.

“But then all the a lot of the wide shots, a lot of like the battle sequences, that's AI because that's what costs all the…”
PJ Ace · 30:30

“So, I think we're going to go hybrid for a while.”
PJ Ace · 31:00

#hybrid production#virtual production#film budgets#actors

Explainer37:30

AI Ads Cost Less, but a Bad One Can Still Be Very Expensive

The hosts and PJ estimate that AI can reduce premium commercial production costs by roughly an order of magnitude, with an entry outsourcing market around $10,000 to $30,000. PJ warns that price alone is the wrong metric for major brands because visibly weak AI work can trigger costly negative publicity.

AI commercial pricing spans entry projects through six-figure campaigns.
A former $3 million production might be approximated at around $300,000.
The cited entry outsourcing range is roughly $10,000 to $30,000.
Campaign scope and strategic work can raise costs substantially.
Reputational damage can outweigh production savings.

“But we always tell brands like what we're delivering is like up against, you know, these other million dollar commercials. And it's like in order…”
PJ Ace · 38:30

“cuz it's far more expensive for you guys to have a negative PR.”
PJ Ace · 39:00

#production costs#brand risk#commercials#pricing

Story· 2

Story01:30

How Kalshi Shipped an NBA Finals Ad in Two Days

PJ Ace explains that Kalshi's fast approval process allowed him to create a culturally specific NBA Finals commercial in roughly two days. AI made it possible to reference the actual teams and moment instead of shooting a generic campaign months before the event.

Fast approvals mattered as much as fast generation.
The commercial could respond to the actual NBA Finals matchup.
Cultural specificity replaced the generic creative required by long production lead times.
Florida characters and Easter eggs gave the ad a distinctive point of view.

“It's in 2 days. We've seen a couple of your viral videos. Just YOLO this thing and like send it to us. Make sure it's…”
PJ Ace · 02:00

“It's It's speed. That's That's the beauty of it.”
PJ Ace · 02:30

#kalshi#speed#cultural relevance#advertising

Story14:30

The VO3 Dialogue Bug That Forced a Character's Death

Kieran recounts how VO3 repeatedly assigned both characters' lines to the human character in his sitcom-style AI ad. His absurd workaround was to declare that the human had died, leaving the computer as the only character able to deliver the remaining dialogue.

VO3 struggled to assign dialogue correctly between a computer and a person.
The model made the human character speak the computer's lines.
Removing one speaking character resolved the attribution problem.
Two human characters might have been easier for the model to distinguish.

“And so the only way I could get this hospital bed scene working is you'll see here that I kill him.”
Kieran Flanagan · 15:00

“So if you look here, this is him dying here.”
Kieran Flanagan · 15:00

#vo3#dialogue#model failure#workaround

Tool· 3

Tool08:00

Why Reference Images Replaced Complicated LoRA Workflows

PJ describes using Ideogram and Google's Nano Banana Pro to place authorized public figures into generated campaign imagery. He contrasts the simple reference-image workflow with older LoRA and ComfyUI processes that demanded far more technical setup.

Ideogram provided access to the image-generation workflow.
Nano Banana Pro could transform a supplied reference image through natural-language instructions.
Older likeness workflows often required custom LoRAs and complex ComfyUI graphs.
Authorized brand partners were enthusiastic about using their likenesses.

“It is now so easy. You can simply add a reference image of of David Beckham and just say, "Okay, now I want him holding…”
PJ Ace · 09:00

#ideogram#nano banana pro#likeness#image generation

Tool17:00

Turn a Clip's Last Frame Into a Clean Close-Up Cut

PJ diagnoses an awkward wide-to-wide transition and recommends extracting the final usable frame from the first clip. The creator can then ask an image model to upscale it and produce a close-up, creating a more natural editorial transition.

Wide-to-wide cuts made the walking sequence feel discontinuous.
A close-up can hide movement and direction inconsistencies.
The last frame of an existing clip preserves scene context.
The extracted frame can become the reference for the next generated shot.

“The issue is you're going from wide to wide, but if you Yeah, if you just do it close up the camera.”
PJ Ace · 17:00

“So you can like play the video through the point you want to cut, take that as a screenshot, and then say, "Hey, upscale this…”
PJ Ace · 17:00

#editing#continuity#close-up#ai video

Tool20:00

Kling Motion Control Lets One Performance Drive Any Character

PJ demonstrates Kling 2.6 motion control, which applies a recorded actor's performance to a generated character image. The technique resembles expensive facial and body performance capture but requires only an image and a driving video.

The workflow combines one generated image with one performance video.
The source performance drives the generated character's movement and expression.
A single performer can portray multiple visually distinct characters.
The setup can produce results in minutes rather than requiring a traditional motion-capture stage.

“You you upload an image and you upload a video and it drives this image with this video's performance.”
PJ Ace · 20:30

“So that person can just play all these different characters.”
Host · 21:00

#kling#motion control#performance capture#characters

Takeaway· 2

Takeaway32:30

Make Spec Ads to Turn AI Video Practice Into Paid Work

PJ advises beginners to follow their curiosity rather than treating creativity as an obligation. For creators seeking commercial opportunities, he identifies speculative advertisements as the clearest wedge because they demonstrate useful capabilities and can lead directly to brand work.

Start with worlds and ideas that naturally sustain curiosity.
Playful experimentation builds familiarity with the tools.
Spec ads can act as a portfolio before a creator has clients.
Commercial work currently offers a clearer income path than unauthorized fan films.
Small in-house teams can also adopt the tools without a large agency.

“just play around and have fun with what drives your curiosity.”
PJ Ace · 32:30

“if you're a young creator, just make crazy spec ads”
PJ Ace · 33:30

#spec ads#creative careers#learning#agency

Takeaway41:00

Turn the Uncanny Valley Into a Creative Constraint

When a client prohibited AI-generated humans, PJ's team built a playful commercial around aliens instead. He recommends shifting toward animals, aliens, or deliberately heightened concepts when photorealistic people would expose current model weaknesses.

A no-human requirement can generate a more distinctive concept.
Aliens and animals reduce sensitivity to imperfect human realism.
Over-the-top ideas fit the current strengths of AI video.
Constraints can improve creativity rather than merely limiting production.

“they were like, "No humans." So we're like, "Okay, we'll do aliens."”
PJ Ace · 41:00

“And it came from the constraint of like we don't want to use humans.”
PJ Ace · 41:00

#uncanny valley#creative constraints#aliens#concepting