Hollywood Insider Spills the SECRETS Behind the AI Video Revolution... | Tim of Theoretically Media

AI is rapidly transforming the world of content creation, from Hollywood to independent studios, with major implications for artists, producers, and audiences. This discussion explores the evolving landscape of AI-generated video, challenging traditional notions of art and authorship, and highlighting the emergence of new creative and business opportunities. It’s a wild ride where the line between AI and reality is blurring faster than anyone expected.

Here’s a breakdown of the key insights and discussions:

• Artist Reactions to AI: There are strong, often polarizing, emotions within the artist community regarding AI. While some artists are very angry, even to the point of suggesting violence against data centers, many opinions likely fall somewhere in the middle.
• Job Displacement vs. Job Creation: AI is seen as potentially eliminating very specific, entry-level, and repetitive jobs, like rotoscoping (cutting people out of photos). However, it's believed that AI will ultimately create more opportunities, enabling smaller teams or even individuals to produce high-quality content.
• The Future of Production Studios: Large, "Orwellian" animation studios with hundreds of animators might shrink significantly. Instead, there will likely be many more smaller production companies, possibly 10-15 people, creating more diverse content for social media and streaming platforms.
• Personal Authorship in AI Creation: When prompting AI, a sense of authorship is still felt, especially when combining AI tools with personal input like shooting clips on a phone. The fun part is seeing what you can do by marrying AI with physical camera work.
• Sky Glass (RIP) Example: The now-defunct Sky Glass company offered a cool service where you could film yourself and it would instantly rotoscope you out and place you in a 3D Unreal background, allowing for virtual sets like a 1940s noir office. Running that through another AI video-to-video filter could produce impressive, albeit "janky," results, offering a cheap alternative to expensive virtual stages like Disney's "The Volume" used for The Mandalorian.
• Rick Rubin and "Vibe Coding": Legendary music producer Rick Rubin is seen as a champion of "vibe coding," which essentially means having great taste to draw out amazing results from artists, even without technical skill. This parallels how people use AI models: you need the taste and vision to prompt them effectively and get the right output.
• Defining Art in the Age of AI: The term "artist" is loaded. It's preferred to think of creators as "producers" or simply "makers" of something when using AI. The definition of art itself is debated – is it art when you pick up crayons, or only when you're paid? Even AI can produce "uninteresting" or "boring" "content."
• The Evolution of Prompt Engineering: The era of "prompt engineering courses" is over. The true creativity in using tools like Midjourney now lies in "remixing" disparate visual ideas and knowing how to describe them to create something new and interesting (e.g., combining Indian, cowboy, and Viking cultures).
• Genre Shifts in AI Content:
  • Thriving Genres: Horror and comedy are predicted to "kill it" in AI video. AI naturally excels at creating "liminal spaces" (eerie, out-of-context environments), which are inherently unsettling and perfect for horror. Comedy benefits from AI's ability to generate silly, absurd situations easily.
  • Dying Genres: Cyberpunk and sci-fi might slowly phase out in AI, as they are "obvious" to generate and audiences are becoming uninterested in repetitive visuals like "rain, neon-soaked streets, and people in black leather with robot eyes."
• Historical Accuracy Challenges: AI struggles with precise historical accuracy, often getting details like uniforms or airplane parts wrong, making it less suitable for sticklers for detail. It's fine for broad overviews, but not for maintaining meticulous historical accuracy.
• Recasting Movies with AI: While there might be initial novelty in seeing a famous movie recast with a new actor (e.g., Happy Gilmore starring Jennifer Aniston), it's unlikely to last. The cost to generate such content for companies would be high, and the novelty might quickly wear off, with audiences returning to the originals.
• The "10:10 Watch Problem": AI image generators, trained on countless watch advertisements, famously display watches at 10:10 regardless of the requested time. Google's Veo 3 tried to get around this by showing an "upside-down" watch, but the 10:10 remains a subtle influence. This highlights AI's current limitations in truly understanding the physics and functions of objects beyond pattern recognition.
• Google's Veo 3 as a "ChatGPT Moment": Veo 3 is seen as a pivotal breakthrough, similar to ChatGPT, because it breaks down barriers to content creation. It combines audio and video, allowing for the production of high-quality content without needing traditional connections, technical knowledge, or massive budgets.
• New Business Opportunities in Video: The removal of barriers by tools like Veo 3 means that companies started by non-media professionals could generate millions within five years. There's a strong potential for businesses to leverage AI video for things like e-commerce, video sales letters, and marketing, as AI can generate compelling copy and visuals quickly and affordably.
• AI in Advertising and Marketing: AI video, especially tools like Veo 3 (once it supports vertical video and product references), is expected to "blow up" in advertising and marketing. Imagine a supermodel interacting with your product via AI-generated video for TikTok or Instagram – the potential for viral campaigns is immense.
• AI Video for Entertainment: AI-generated videos are starting to be watched for pure entertainment value, seamlessly blending into feeds. The "plastic baby" video by Meta Puppet is highlighted as an example of compelling AI-generated content that effectively shifts scenarios and maintains audience engagement.
• The "Uncanny Valley" and AI Horror: Old "janky" AI video had a "cursed" appeal, with wrong physics and melting people, akin to "America's Funniest Home Videos." With tools like Veo 3, the quality is improving, leading to "correct" and genuinely funny or engaging AI clips, particularly in horror, where AI can naturally create scary and unsettling visuals.
• Stand-up Comedy and Reddit Influence: AI-generated stand-up comedy is gaining traction, possibly due to Google training models on Reddit data, which is known for its "snarky" and "dad jokes" style humor.
• Long-Form Narrative and Editing: For creating long-form narratives with AI video, traditional editing tools like DaVinci Resolve or Premiere Pro are still essential. AI primarily generates "raw material," and the true skill lies in editing, filtering, and managing consistency, physics, and other random quirks of AI-generated video.
• The Speed of AI Advancement: The pace of AI development is far exceeding expectations. What was predicted to be four years away in 2022 is now here, with the "goalpost" for "there yet" constantly moving.
• The "Terminus Point" for AI Video: It's unclear when AI video will be considered "done" or "there" for everyone. Is it when it looks like a low-budget Netflix show, or a high-budget action film? The real question is whether people want to actively engage with AI video to "prompt the latest Mission Impossible movie" or if something else matters more.
• The Quest for "Soul" in AI: Some feel AI video still lacks "soul" or "emotion." While it can be "pretty," it's not yet "beautiful" or "complete" in a way that resonates deeply, particularly in the subtleties of human reactions.
• AI for IP Creation: The current era is a golden opportunity to create new Intellectual Property (IP) instead of constantly chasing existing, corporate-owned franchises like Batman or Star Wars. With AI, you have "infinite imagination" to create unique, original content. If it becomes a viral hit, studios will come to you, allowing for better negotiation terms than what original comic creators often received.
• YouTube as a Distribution Pipeline: For new IP, the suggested pipeline is to create your own thing, release it on YouTube, and if it gains traction, distribution platforms like Amazon Prime (as seen with the Murder Drones anime) or even Netflix might pick it up.
• Netflix's Internal AI Animation Studio: Netflix is known to be brewing its own AI animation studio, likely focusing on shorter, experimental content similar to Love, Death & Robots, rather than giant shows. This means independent artists with AI tools are increasingly on a level playing field with major studios.
• Fable Studios' Showrunner and Thistle Gulch:
  • Showrunner: This project aimed for "infinite storytelling" using AI to replicate the South Park visual style (due to its simplicity). It used Stable Diffusion for visuals, Chat GPT for dialogue, and ElevenLabs for voices. It simulates an entire town with characters (like Stan and Kyle) having their own motivations and conversations, with the AI stringing together interesting moments into a "movie."
  • Thistle Gulch: A more fascinating, but less developed, project was a 3D game engine of a Western town. It featured AI agents (characters) with their own motivations, moving around in a murder mystery, allowing a free-roaming camera and user influence. This hints at the "holodeck idea" – immersive VR experiences with AI agents.
• Passive vs. Interactive Entertainment: While interactive content (like Netflix's Bandersnatch) has had limited success, the future might see a split. Some people will still prefer passive viewing – just zoning out and watching TV without having to make choices or interact. Others will embrace active, agentic TV or VR experiences where they can influence storylines and explore virtual worlds.
• AI for Video Games and VR: The deep simulation capabilities of AI, while potentially "overkill" for linear movies, are considered "absolutely incredible" for video games and VR. The "holodeck aspect" of AI-driven virtual worlds is seen as a new branch of entertainment, not necessarily replacing filmmaking, but offering a distinct and engaging experience.