StarZero All shows Try the podcast clipper →

May 28, 2025 · 46m · startup-ideas

How I use Google Veo3 to create viral videos (3M views in 48 hrs)

PJ Accetturo · 33m spoken Greg Isenberg · 7m spoken Commercial Character (Puppermin) · 33s spoken
0:00 / 0:00
▶ Watch on YouTube →

gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions

Host Greg Isenberg interviews viral creator PJ to break down the step-by-step workflow of scripting, prompting, and editing Google Veo 3 videos, alongside actionable strategies for monetizing AI video production.

How this conversation actually went

Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. Greg holds 19% of the talking time here. How this is scored →

Greg as informed peer 4.9 Guest teaching 6.0 Guest disagreement 1.6 Greg pushing back 1.1
05100:0015:0030:0045:000:50–4:00 · Greg as informed peer 3/10 Demystifying AI Video Production and Niche Opportunities Greg sets up the premise of the episode, asking whether everyday creators need technical chops to leverage Veo 3. PJ explains how LLMs simplify cinematic generation and highlights the importance of targeting niche communities.4:01–12:47 · Greg as informed peer 5/10 Analyzing Viral Engagement and Content Psychology PJ shares his screen to demonstrate how he iteratively brainstorms satirical commercial ideas in ChatGPT. Greg actively contributes by noting how users can benchmark satire using established models like SNL and Tim and Eric.12:48–16:33 · Greg as informed peer 5/10 Iterative Joke Selection and Script Refinement PJ outlines the 95/5 rule of AI comedy generation, explaining how to harvest rare winning one-liners and refeed them into the model. Greg adds that creators lacking comedic taste can outsource style by instructing the model to emulate specific comedians like Seth Rogen.16:33–21:16 · Greg as informed peer 3/10 Shot-by-Shot Prompting and Google Veo 3 Execution PJ educates Greg on the operational quirks of Google Veo 3, including its unpredictable audio generation, subtitle glitches caused by quotation marks, and current character consistency limitations. Greg listens and probes about production costs and audio bugs.21:16–25:21 · Greg as informed peer 4/10 Timeline Editing, Camera Motion, and Audio Nuances PJ walks through his Final Cut Pro timeline, demonstrating how detailed emotional and camera movement prompting (such as dolly-ins) creates non-AI realism. Greg clarifies editor choices and audio workflow.25:22–30:15 · Greg as informed peer 8/10 Crafting High-Converting Hooks and Distribution on X Greg displays strong domain mastery by deconstructing the psychological and algorithmic mechanics of PJ's viral tweet hook. PJ enthusiastically validates Greg's analysis and shows further headline examples.30:15–44:45 · Greg as informed peer 6/10 Ten Actionable Monetization Strategies for AI Creators PJ details ten monetization pathways for generative video creators while expressing strong skepticism toward course-selling and traditional Hollywood-style feature films. Greg pushes back gently on the course critique by reframing it as high-value digital product bundling.0:50–4:00 · Guest teaching 5/10 Demystifying AI Video Production and Niche Opportunities Greg sets up the premise of the episode, asking whether everyday creators need technical chops to leverage Veo 3. PJ explains how LLMs simplify cinematic generation and highlights the importance of targeting niche communities.4:01–12:47 · Guest teaching 6/10 Analyzing Viral Engagement and Content Psychology PJ shares his screen to demonstrate how he iteratively brainstorms satirical commercial ideas in ChatGPT. Greg actively contributes by noting how users can benchmark satire using established models like SNL and Tim and Eric.12:48–16:33 · Guest teaching 6/10 Iterative Joke Selection and Script Refinement PJ outlines the 95/5 rule of AI comedy generation, explaining how to harvest rare winning one-liners and refeed them into the model. Greg adds that creators lacking comedic taste can outsource style by instructing the model to emulate specific comedians like Seth Rogen.16:33–21:16 · Guest teaching 7/10 Shot-by-Shot Prompting and Google Veo 3 Execution PJ educates Greg on the operational quirks of Google Veo 3, including its unpredictable audio generation, subtitle glitches caused by quotation marks, and current character consistency limitations. Greg listens and probes about production costs and audio bugs.21:16–25:21 · Guest teaching 6/10 Timeline Editing, Camera Motion, and Audio Nuances PJ walks through his Final Cut Pro timeline, demonstrating how detailed emotional and camera movement prompting (such as dolly-ins) creates non-AI realism. Greg clarifies editor choices and audio workflow.25:22–30:15 · Guest teaching 5/10 Crafting High-Converting Hooks and Distribution on X Greg displays strong domain mastery by deconstructing the psychological and algorithmic mechanics of PJ's viral tweet hook. PJ enthusiastically validates Greg's analysis and shows further headline examples.30:15–44:45 · Guest teaching 7/10 Ten Actionable Monetization Strategies for AI Creators PJ details ten monetization pathways for generative video creators while expressing strong skepticism toward course-selling and traditional Hollywood-style feature films. Greg pushes back gently on the course critique by reframing it as high-value digital product bundling.0:50–4:00 · Guest disagreement 1/10 Demystifying AI Video Production and Niche Opportunities Greg sets up the premise of the episode, asking whether everyday creators need technical chops to leverage Veo 3. PJ explains how LLMs simplify cinematic generation and highlights the importance of targeting niche communities.4:01–12:47 · Guest disagreement 1/10 Analyzing Viral Engagement and Content Psychology PJ shares his screen to demonstrate how he iteratively brainstorms satirical commercial ideas in ChatGPT. Greg actively contributes by noting how users can benchmark satire using established models like SNL and Tim and Eric.12:48–16:33 · Guest disagreement 1/10 Iterative Joke Selection and Script Refinement PJ outlines the 95/5 rule of AI comedy generation, explaining how to harvest rare winning one-liners and refeed them into the model. Greg adds that creators lacking comedic taste can outsource style by instructing the model to emulate specific comedians like Seth Rogen.16:33–21:16 · Guest disagreement 2/10 Shot-by-Shot Prompting and Google Veo 3 Execution PJ educates Greg on the operational quirks of Google Veo 3, including its unpredictable audio generation, subtitle glitches caused by quotation marks, and current character consistency limitations. Greg listens and probes about production costs and audio bugs.21:16–25:21 · Guest disagreement 1/10 Timeline Editing, Camera Motion, and Audio Nuances PJ walks through his Final Cut Pro timeline, demonstrating how detailed emotional and camera movement prompting (such as dolly-ins) creates non-AI realism. Greg clarifies editor choices and audio workflow.25:22–30:15 · Guest disagreement 2/10 Crafting High-Converting Hooks and Distribution on X Greg displays strong domain mastery by deconstructing the psychological and algorithmic mechanics of PJ's viral tweet hook. PJ enthusiastically validates Greg's analysis and shows further headline examples.30:15–44:45 · Guest disagreement 3/10 Ten Actionable Monetization Strategies for AI Creators PJ details ten monetization pathways for generative video creators while expressing strong skepticism toward course-selling and traditional Hollywood-style feature films. Greg pushes back gently on the course critique by reframing it as high-value digital product bundling.0:50–4:00 · Greg pushing back 1/10 Demystifying AI Video Production and Niche Opportunities Greg sets up the premise of the episode, asking whether everyday creators need technical chops to leverage Veo 3. PJ explains how LLMs simplify cinematic generation and highlights the importance of targeting niche communities.4:01–12:47 · Greg pushing back 1/10 Analyzing Viral Engagement and Content Psychology PJ shares his screen to demonstrate how he iteratively brainstorms satirical commercial ideas in ChatGPT. Greg actively contributes by noting how users can benchmark satire using established models like SNL and Tim and Eric.12:48–16:33 · Greg pushing back 1/10 Iterative Joke Selection and Script Refinement PJ outlines the 95/5 rule of AI comedy generation, explaining how to harvest rare winning one-liners and refeed them into the model. Greg adds that creators lacking comedic taste can outsource style by instructing the model to emulate specific comedians like Seth Rogen.16:33–21:16 · Greg pushing back 1/10 Shot-by-Shot Prompting and Google Veo 3 Execution PJ educates Greg on the operational quirks of Google Veo 3, including its unpredictable audio generation, subtitle glitches caused by quotation marks, and current character consistency limitations. Greg listens and probes about production costs and audio bugs.21:16–25:21 · Greg pushing back 1/10 Timeline Editing, Camera Motion, and Audio Nuances PJ walks through his Final Cut Pro timeline, demonstrating how detailed emotional and camera movement prompting (such as dolly-ins) creates non-AI realism. Greg clarifies editor choices and audio workflow.25:22–30:15 · Greg pushing back 1/10 Crafting High-Converting Hooks and Distribution on X Greg displays strong domain mastery by deconstructing the psychological and algorithmic mechanics of PJ's viral tweet hook. PJ enthusiastically validates Greg's analysis and shows further headline examples.30:15–44:45 · Greg pushing back 2/10 Ten Actionable Monetization Strategies for AI Creators PJ details ten monetization pathways for generative video creators while expressing strong skepticism toward course-selling and traditional Hollywood-style feature films. Greg pushes back gently on the course critique by reframing it as high-value digital product bundling.

speaking balance: gold is Greg, purple is the guest (3 minute bins)

0:00 · Greg 61.2% · guest 38.8%0:00 · Greg 61.2% · guest 38.8%3:00 · Greg 18.1% · guest 81.9%3:00 · Greg 18.1% · guest 81.9%6:00 · Greg 11.7% · guest 88.3%6:00 · Greg 11.7% · guest 88.3%9:00 · Greg 17.2% · guest 82.8%9:00 · Greg 17.2% · guest 82.8%12:00 · Greg 9.8% · guest 90.2%12:00 · Greg 9.8% · guest 90.2%15:00 · Greg 7.4% · guest 92.6%15:00 · Greg 7.4% · guest 92.6%18:00 · Greg 4.1% · guest 95.9%18:00 · Greg 4.1% · guest 95.9%21:00 · Greg 9.8% · guest 90.2%21:00 · Greg 9.8% · guest 90.2%24:00 · Greg 21% · guest 79%24:00 · Greg 21% · guest 79%27:00 · Greg 32% · guest 68%27:00 · Greg 32% · guest 68%30:00 · Greg 37.6% · guest 62.4%30:00 · Greg 37.6% · guest 62.4%33:00 · Greg 0% · guest 100%33:00 · Greg 0% · guest 100%36:00 · Greg 3.5% · guest 96.5%36:00 · Greg 3.5% · guest 96.5%39:00 · Greg 23.2% · guest 76.8%39:00 · Greg 23.2% · guest 76.8%42:00 · Greg 7.8% · guest 92.2%42:00 · Greg 7.8% · guest 92.2%45:00 · Greg 63.6% · guest 36.4%45:00 · Greg 63.6% · guest 36.4%
Sharpest disagreement ▶ 41:20 Rejecting the traditional Hollywood long-form model for AI video

PJ forcefully dismisses the viability of producing 90-minute AI feature films, arguing that rapid iteration makes long-form models obsolete compared to episodic weekly UGC.

 Hardest push from Greg ▶ 39:53 Reframing course selling as digital product packaging

When PJ mocks selling marketing courses as an MLM scheme, Greg immediately interjects to reframe the concept into packaging actionable digital products and PDFs.

 Biggest teaching moment ▶ 18:43 Explaining Veo 3 generation bugs and quote-formatting tricks

PJ details key undocumented technical nuances of Veo 3, teaching Greg why quotes trigger gibberish subtitles and how the platform's audio generation acts like an unpredictable slot machine.

 Greg holds their own ▶ 27:02 Deconstructing viral copywriting and engagement triggers

Greg takes command of the conversation by systematically breaking down why PJ's post worked, citing price-point symmetry, subtle credibility flexing, and question framing designed for algorithmic distribution.
the scores for every segment, with the reasoning behind each
ChapterTopicGreg as informed peerGuest teachingGuest disagreementGreg pushing backWhy
Demystifying AI Video Production and Niche Opportunities 3511 Greg sets up the premise of the episode, asking whether everyday creators need technical chops to leverage Veo 3. PJ explains how LLMs simplify cinematic generation and highlights the importance of targeting niche communities.
Analyzing Viral Engagement and Content Psychology 5611 PJ shares his screen to demonstrate how he iteratively brainstorms satirical commercial ideas in ChatGPT. Greg actively contributes by noting how users can benchmark satire using established models like SNL and Tim and Eric.
Iterative Joke Selection and Script Refinement 5611 PJ outlines the 95/5 rule of AI comedy generation, explaining how to harvest rare winning one-liners and refeed them into the model. Greg adds that creators lacking comedic taste can outsource style by instructing the model to emulate specific comedians like Seth Rogen.
Shot-by-Shot Prompting and Google Veo 3 Execution 3721 PJ educates Greg on the operational quirks of Google Veo 3, including its unpredictable audio generation, subtitle glitches caused by quotation marks, and current character consistency limitations. Greg listens and probes about production costs and audio bugs.
Timeline Editing, Camera Motion, and Audio Nuances 4611 PJ walks through his Final Cut Pro timeline, demonstrating how detailed emotional and camera movement prompting (such as dolly-ins) creates non-AI realism. Greg clarifies editor choices and audio workflow.
Crafting High-Converting Hooks and Distribution on X 8521 Greg displays strong domain mastery by deconstructing the psychological and algorithmic mechanics of PJ's viral tweet hook. PJ enthusiastically validates Greg's analysis and shows further headline examples.
Ten Actionable Monetization Strategies for AI Creators 6732 PJ details ten monetization pathways for generative video creators while expressing strong skepticism toward course-selling and traditional Hollywood-style feature films. Greg pushes back gently on the course critique by reframing it as high-value digital product bundling.

Statements from this episode (2)

Insight
Creators Can Outsource Comedic Taste by Prompting Specific Comedian Personas
“I think one thing you can do If you don't have the taste, you can kind of outsource taste, and you could say, you know, let's say you like Seth Rogen. You'd say, like, create jokes that I think Seth Rogen would like.”
Greg Isenberg May 28, 2025 ▶ 15:35
Insight
AI Video Brings Cinematic Commercials to Niches Priced Out of Agencies
“This opens the door to a huge number two basically creating commercials for different niches. This opens the door to So many different like vertical niches that wouldn't have this level of cinematography and couldn't afford a 500,000 dollar agency for a commer…”
Greg Isenberg May 28, 2025 ▶ 45:13
Made with StarZero

Turn any episode into a week of clips.

This entire site, over 300 episodes transcribed, diarized, checked and made playable, runs on the StarZero media pipeline. Drop in your own episode and the podcast clipper finds the moments worth sharing, cuts them, captions them, and reframes them for every feed.

Try the podcast clipper Learn more