10 AI Video Trends Taking Over the Internet
10 AI Video Trends Taking Over the Internet
Overview
This episode of the AI Daily Brief (dated June 21, 2025) covers two segments. The headlines segment recaps major AI industry news, including Sam Altman’s confirmation of Meta’s nine-figure talent recruitment offers, OpenAI’s first Pentagon contract, and Cursor’s $200/month Ultra Plan launch. The main episode focuses on the central thesis: Google’s VO3 video generation model represents a genuine inflection point in AI-generated content, having spawned an entirely new ecosystem of viral video formats across social media platforms. The speaker argues that VO3’s key unlock — native, prompt-aligned audio generation alongside video — fundamentally changes what AI video creation can produce and how it spreads. No individual speaker name or affiliation is explicitly stated beyond the show title.
Source video URL: Not provided.
Prerequisites
- Familiarity with generative AI concepts (text-to-image, text-to-video models)
- Basic understanding of social media content ecosystems (TikTok, Instagram Reels, YouTube Shorts)
- Awareness of prior AI video/image tools (Midjourney, Sora, etc.)
- General knowledge of the competitive AI lab landscape (OpenAI, Meta, Google, Anthropic)
- Understanding of ASMR as a content genre
Main Points
Headlines: Meta’s Nine-Figure Talent War
- Sam Altman, appearing on his brother Jack Altman’s podcast, confirmed that Meta/Mark Zuckerberg is offering $100 million+ signing bonuses and more than that in annual compensation to poach OpenAI researchers.
- Just one month prior, reports had described $20 million packages as the extreme end; the scale of offers has escalated dramatically in a short window.
- Noam Brown (leading reasoning and agentics researcher) was approached and declined; he had previously left Meta for OpenAI in 2023 citing mission alignment over maximum compensation.
- Altman argued that leading with massive upfront compensation rather than mission creates poor culture, and that OpenAI staff believe they have the better shot at delivering superintelligence.
- Strategically, Altman’s public confirmation frames anyone who accepts Meta’s offers as a mercenary, while casting doubt among Meta employees who were not offered comparable packages.
Headlines: OpenAI’s First Pentagon Contract ($200M)
- OpenAI secured a $200 million Pentagon contract under a new entity called OpenAI for Government, consolidating prior partnerships (NASA, U.S. National Labs, U.S. Treasury).
- Stated use cases include healthcare access for service members, acquisition data analysis, and proactive cyber defense.
- The Pentagon’s own language was broader: developing “frontier AI capabilities to address critical national security challenges in both warfighting and enterprise domains.”
- This follows OpenAI removing its total ban on military usage in January 2024 and subsequently partnering with Anduril on anti-drone targeting.
- The contract may undercut Microsoft’s existing federal government cloud contracts, as Microsoft had only recently received approval for its hosted OpenAI service at all classification levels.
Headlines: The $200/Month AI Subscription Tier Expands
- Cursor launched a $200/month Ultra Plan offering 20x more usage than its Pro tier, targeting power users who want predictable pricing over usage-based billing.
- The $20/month plan was also upgraded to unlimited access subject to rate limits.
- This follows OpenAI and Anthropic both establishing $200/month tiers; Cursor is the first AI tool (rather than model provider) to experiment at this price point.
- Separately, Cursor is reportedly fielding acquisition/investment interest at an $18–20 billion valuation, just two weeks after closing a Series C at $9.9 billion.
- Cursor has reached $500 million ARR, a 60% increase in two months, described as the fastest-growing startup in Silicon Valley history.
Main Episode: VO3 as an Inflection Point
- Google’s VO3 model, released weeks before this episode, has rapidly dominated social media feeds on Instagram Reels and TikTok.
- The critical differentiator is native, synchronized audio generation within a single prompt — characters speak, sounds occur, and video matches, all generated together.
- Previous workflows required generating video and audio separately, then manually synchronizing them; VO3 collapses this into one step.
- YouTube CEO Neil Mohan announced VO3 integration directly into YouTube Shorts (revealed at the Cannes Lions advertising event); YouTube Shorts average 200 billion views per day.
- Midjourney also released its video model the same week, praised for its distinctive aesthetic and visual legacy, but it lacks VO3’s native audio capability.
Trend 1: ASMR Videos
- AI-generated glass fruit-cutting ASMR videos are going viral (one cited example: 2.5 million views, 610,000 likes).
- VO3’s integrated sound is described as having “digested the motherlode of ASMR content on YouTube,” making it a natural AI ASMR generator.
- One example reached 3.1 million likes and 12,000 comments in three days.
Trend 2: Absurdist/Dumb Comedy (Sinkhole Videos)
- Short, deadpan comedy clips — e.g., a newscaster reporting live as people fall into sinkholes — are a popular emerging format.
- Simple premise, quick punchline; designed for short-form virality.
Trend 3: Bigfoot/Yeti Vlogs
- Possibly the dominant VO3 genre since the model’s release; hundreds to thousands of channels producing content.
- Format: Bigfoot or Yeti characters narrate daily life in vlog style, adopting social media creator tropes (energy drinks, trail cams, etc.).
- Content ranges from PG-13 to NSFW; broadly described as “hilarious” and “genuinely funny” even by self-described AI skeptics.
Trend 4: Historical Figures and IP Characters as Social Media Creators
- The vlog-style format extends to historical figures (e.g., Revolutionary War–era Boston) and biblical characters as influencers.
- Creator PJ Ace’s Bible influencer videos went viral and led to a nationally aired NBA Finals ad for Calci, described as “the most unhinged ad that’s ever aired on national TV.”
- PJ Ace, a director with 15+ years of experience, noted: “Just because this was cheap doesn’t mean anyone can do it. Brands still pay a premium for taste.”
Trend 5: Fan-Made IP Spin-offs (Harry Potter, Star Wars)
- VO3 is enabling a new category of fan-made spin-off content using well-known IPs.
- Vlogwarts (Harry Potter vlog): 115,000 followers, 2 million likes from 10 videos in ~12 days.
- Stormtrooper Vlogs (Star Wars): 0 to 300,000 followers after 20 videos.
- Content evolved from retelling source material to original, increasingly “unhinged” crossover narratives (e.g., Harry Potter meets The Hangover).
- Commentator Trung Phan argued this signals significant future impact on Hollywood, predicting that IP value, curation, distribution, community building, and marketing expertise will all increase in importance.
- Even typically anti-AI Reddit communities showed mixed-to-positive reactions, with comments like “AI sucks, but this is funny AF.”
- Key insight: creators are not one-shotting these videos; craft, character consistency, and narrative decisions still differentiate quality content.
Trend 6: Small Business Advertising
- Small business owners are adapting viral VO3 formats (especially Bigfoot vlogs) for local advertising.
- A West Hollywood dentist’s Bigfoot vlog ad went viral.
- A plastic surgeon’s parody adventure ad similarly circulated widely.
- A16Z partner Justine Moore highlighted this as a potentially dominant force in small business marketing.
Trend 7: Celebrity Educational Videos
- Accounts like Unlock Learning (active for 18 months) have reached a new level using VO3, featuring AI-generated celebrity avatars teaching math and other subjects.
- Viewer comments show genuine engagement and learning.
- Compared to Khan Academy’s whiteboard format — same underlying mechanism (relatable, informal explanation), new toolset.
Trend 8: Fantasy and Sci-Fi Worldbuilding
- More elaborate, high-fidelity fantasy narratives (e.g., Titan-class organism lore, Atlantis series by Gossip Goblin) are gaining traction beyond quick memes.
- Content extends to AI horror, including atmospheric short films set in isolated locations with creeping dread.
- “Point of view: you wake up as [mythological figure]” is a separate sub-genre with tens of millions of views.
Trend 9: Meta-Commentary and Mockumentary
- Creators are using VO3 to make mockumentaries about AI itself (e.g., The Prompt Floor, a behind-the-scenes look at how AI videos are made from the perspective of the generated characters).
- Cited as evidence that creativity and expanded narrative ambition are growing alongside technical capability.
Trend 10: Established Comedians Transitioning to AI Video
- Comedian John LaJoy, with a decade-long YouTube background, has successfully transitioned to AI-generated TikTok content.
- His format (e.g., Jesus hosting a podcast, robots from the future recording a podcast warning humanity) combines traditional comedy writing with VO3 generation.
- Demonstrates that existing creative talent can leverage these tools as a new medium rather than a replacement for craft.
Key Concepts
- VO3 (Google Veo 3): Google’s text-to-video generation model that natively produces synchronized audio and video from a single prompt, released in mid-2025.
- Midjourney Video Model: Midjourney’s long-anticipated entry into video generation, praised for its distinctive aesthetic legacy but lacking native audio at launch.
- ASMR (Autonomous Sensory Meridian Response): A content genre characterized by soft sounds and visuals designed to produce a relaxing, tingly sensation in viewers.
- Vlog-style AI video: A format applying social media creator tropes (first-person narration, casual speech, everyday scenarios) to fictional or non-human characters.
- Fan fiction spin-off content: AI-generated narratives using established IP characters (Star Wars, Harry Potter) in new, original scenarios without studio involvement.
- Greg the Stormtrooper: A recurring everyman character in the Stormtrooper Vlogs account, cited as an example of how consistent character identity elevates AI-generated content.
- OpenAI for Government: A new OpenAI entity consolidating federal government partnerships, including the inaugural $200M Pentagon contract.
- Cursor Ultra Plan: A $200/month coding AI subscription tier offering 20x usage over the standard Pro plan, the first such tier from an AI tool (non-model) company.
- ARR (Annual Recurring Revenue): A measure of predictable, recurring subscription revenue; Cursor reached $500M ARR as of this episode.
- Nine-figure compensation packages: Offers reportedly exceeding $100M in signing bonuses plus annual compensation, used by Meta to recruit top AI researchers.
- IP (Intellectual Property): Legal ownership of creative works; a central tension in AI-generated fan content using studio-owned characters.
Summary
The episode argues that Google’s VO3 — specifically its ability to generate native, synchronized audio and video from a single prompt — constitutes a genuine inflection point in AI-generated content, having spawned at least ten distinct viral video categories within weeks of its release: ASMR, absurdist comedy, Bigfoot/Yeti vlogs, historical and biblical figure influencer content, fan-made IP spin-offs, small business advertising, celebrity educational videos, fantasy and horror worldbuilding, AI meta-commentary, and established comedians adopting the medium. The speaker emphasizes that while the tooling lowers the barrier to creation, craft, taste, and consistent character-building still differentiate high-performing content from noise — a point reinforced by professional creators like PJ Ace and the anonymous Stormtrooper Vlogs operator. Surrounding this main thesis, the headlines establish the broader competitive and commercial context: a talent war in which Meta is offering nine-figure packages to poach OpenAI researchers, OpenAI deepening its government and defense relationships with a $200M Pentagon contract, and the $200/month premium subscription model spreading from model providers to AI tools like Cursor. Taken together, the episode presents a picture of AI rapidly moving from laboratory novelty to a pervasive force reshaping content creation, advertising, entertainment, and the economics of the AI industry itself.