Jul 9, 2025 · 49m · latent-space
AI Video Is Eating The World — Olivia and Justine Moore, a16z
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
In this episode of the Latent Space podcast, a16z investment partners Justine and Olivia Moore explore the explosive rise of AI video, dissecting viral social media phenomena, creator economics, and automated production pipelines.
How this conversation actually went
Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. How this is scored →
The hosts as informed peer 4.0 Guest teaching 5.0 Guest disagreement 1.6 The hosts pushing back 1.9
speaking balance: gold is the hosts, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 38:42 Challenging the native-versus-repurposed content rule
Justine directly refutes Wix's rule that repurposed podcast clips cannot go viral, pointing to creator accounts like Vitrupo that dominate the algorithm through long-form clipping.
Hardest push from the hosts ▶ 39:19 Refusing clickbait clipping tacticsWix pushes back against purely viral clipping strategies, emphasizing that Latent Space must protect its technical credibility and brand from sensationalized, out-of-context hooks.
Biggest teaching moment ▶ 34:25 Demonstrating automated clipping agent workflowsJustine and Olivia show Wix a live demonstration of automated AI clipping and virality scoring on Overlap, showing him an end-to-end publishing pipeline he was seeking but did not know existed.
The host holds their own ▶ 27:37 Presenting State of AI Engineering survey dataWix counters the perception that Adobe's clean models lack real traction by citing hard survey data showing Adobe outperforming popular dedicated generative model competitors.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | The hosts as informed peer | Guest teaching | Guest disagreement | The hosts pushing back | Why |
|---|---|---|---|---|---|---|
| The Shift to AI Video and the Rise of Italian Brainrot | 2 | 5 | 1 | 1 | The hosts adopt a curious, non-expert stance as Wix admits he does not follow short-form AI video trends. Justine and Olivia explain the organic origins and mechanics of the decentralized Italian brainrot meme universe. | |
| Tracking Virality and Olivia's Veo 3 Experiments | 3 | 6 | 1 | 1 | Olivia shares screen recordings of her hands-on Veo 3 ASMR and trend experiments, while Justine breaks down technical nuances such as Veo 3 falling back to Veo 2 when an initial image frame is supplied. | |
| Character Consistency, Narrative Storytelling, and IP Remixing | 4 | 5 | 1 | 1 | Alessio asks whether known IP familiarity is merely an initial crutch before fading out. Justine provides insights from portfolio A/B testing on how IP recognition combined with surreal twists captures viewer retention. | |
| Virtual Influencers and AI Video Monetization Dynamics | 4 | 5 | 2 | 2 | The hosts and guests explore monetization models including virtual influencers, consulting, and platform creator payouts. Olivia points out the high compute generation costs that squeeze creator margins. | |
| Market Architecture: Foundation Models vs. Enablement Layers | 6 | 4 | 2 | 3 | Wix cites his proprietary State of AI Engineering survey to challenge assumptions about Adobe Firefly's market adoption. Justine acknowledges this by distinguishing between generative fill in-painting utility and net-new video generation. | |
| Case Study: AI Video Pipelines and Automated Clipping Tools | 5 | 6 | 3 | 4 | Justine and Olivia demonstrate the Overlap clipping agent on a16z's YouTube channel to show the hosts how to automate short-form distribution. Wix pushes back against generic clipping and clickbait, but Justine argues repurposed video is currently dominating platform algorithms. | |
| Prompt Theory, AI Merchandising, and Future Outlook | 4 | 4 | 1 | 1 | The discussion covers prompt theory, physical merchandising of AI memes like Bread Klimp, and professional commercial workflows. Alessio closes by pitching an automated merchandise generation startup idea. |
Statements from this episode (16)
Olivia Moore: 90% of TikTok and Reels feeds are AI-generated video
“I mean, if you've been on TikTok or Reels or YouTube Shorts recently, in the past week, probably 90% of your feed is AI generated video.”
Olivia Moore: Hundreds of thousands are making and publishing AI video
“And now there's like, I would guess hundreds of thousands of people making and publishing AI video”
Justine Moore: Viral AI video content now originates on TikTok and Instagram
“Now, especially after VO three and, like,
Like Minimax too, and like with the animals diving and that sort of thing, we've actually seen like a flip where most of the viral content is now originating on the true consumer, like everyday consumer platforms, whic…”
Olivia Moore: Google Veo 3 restricts real people but permits cartoon IP
“The interesting thing about VO three is there's no IP restrictions, at least on cartoon characters.
Yeah, on real people, there are.
On cartoon characters, there are not.”
Justine Moore: Veo 3 lacks image-to-video with audio, reverting to Veo 2
“Cause the thing to know about VO three that I think a lot of people don't know until you use it is you actually can't do image to video with audio.
Google hasn't released that yet.
I would assume for trust and safety reasons in the interface, you can start wit…”
Justine Moore: Video creators rely on recognizable archetypes for character consistency
“Which means that it makes it super hard for, like, character consistency, right, if you're not able to start with an image of the same character over and over again, which is part of why you're seeing so many of these viral trends using things like a stormtroo…”
Justine Moore: Novel native IP will increasingly emerge in AI video
“And so I think over time we'll see more AI, like, IP like that.”
Olivia Moore: AI TikTok character Kim the Gorilla amassed 300K followers
“And you can see, like, all of her videos get hundreds of thousands of likes. She already has, like, 300,000 followers in, like, a really short period of time.”
Justine Moore: AI avatars let creators bypass social media beauty standards
“Before, honestly, to be, like, an Instagram or YouTube influencer, most of them are hot people. And now it's, like, anyone can be a popular influencer, and you don't have to be a hot person. Like, how many friends do you have that have, like, are, like, really…”
Olivia Moore: AI influencer monetization will explode with AI video generation
“We already saw this with whole startups or stacks of products, and Justine looked at a bunch of these, where it was, like, basically make money by building your own Instagram influencer off of AI images. And a lot of people were making, yeah, tens of thousands…”
Olivia Moore: High model costs and rerolls hurt AI video ROI
“My learning of making a bunch of these videos and trying to post them on social feeds is it's actually like very expensive still to produce just because VO three is so expensive. And especially if you're making more complex content than like someone slicing a …”
Justine Moore: Creators choose Fal and Replicate over Google's $125 Veo plan
“What we see is, like, a ton of creators are just going to the model enablement layer, whether it's a consumer-facing interface, like CREA, or more of a developer-facing interface, like a FALL or a Replicate, where you can generate Like, videos on a one-off bas…”
Olivia Moore: Google Flow defaults to Veo 2 and fails on mobile
“Flow and VO, it's, like, VO two is actually the default, and you have to, like, navigate a bunch of, like, little hidden buttons to change it to VO three. So there's, like, YouTube videos with, like, millions of hits of just, like, how do I find VO three withi…”
Justine Moore: Foundation models are eating away ComfyUI use cases
“I would agree that more of the Comfy UI use cases are starting to be eaten away by kind of the core foundation model companies.”
Justine Moore: X algorithm currently drives massive reach for video clips
“I also used to think that clips would never do well, but I feel like, especially in the X algorithm right now, it's like a clipping era. Everyone, for some reason, video is, like, doing super well in the algorithm, and especially if you're, it's a video of, li…”
Olivia Moore: AI creators have a 1–2 day window for platform content arbitrage
“There's almost this, like, content arbitrage thing that happens right now, where something will go viral on one platform, and then there's, like, a one- or two-day window to be, like, the person to post it on the next platform, but it's kind of fascinating to …”