In Response to Controversy Over Fake Movie Trailer Channels Being Monetized By Hollywood Studios Through Content ID, YouTube Demonetizes Channels

⚡️ Headlines

🤖 AI

Google DeepMind urges AGI safeguards ahead of rapid AI evolution - In a newly released paper, DeepMind outlines responsible development strategies for Artificial General Intelligence, emphasizing preparedness for emerging risks. [Axios]

DeepMind details responsible AGI roadmap in major policy blog - DeepMind’s official blog elaborates on its AGI safety plan, focusing on societal alignment, risk mitigation, and global cooperation as AI capabilities advance. [DeepMind Blog]

Alibaba to launch upgraded AI model Qwen 3 this month - Alibaba is preparing to release Qwen 3, a more powerful version of its flagship AI model, as early as April to keep pace with global competitors. [Bloomberg]

Gladia debuts multilingual AI transcription model Solaria - Startup Gladia unveils Solaria, a speech-to-text model supporting 100 languages and real-time transcription across diverse use cases. [VentureBeat]

NaNoWriMo shuts down following AI moderation scandals - National Novel Writing Month ceases operations after controversy over its use of AI tools to moderate content and alleged censorship. [TechCrunch]

Runway’s Gen-4 model brings next-gen AI video to filmmakers - Runway introduces Gen-4, its most advanced video generation model yet, offering multi-shot continuity and improved realism for AI-driven filmmaking. [The Verge]

SoftBank seeks $16.5 billion loan to fund U.S. AI ventures - SoftBank is pursuing a record-setting loan to finance large-scale AI infrastructure projects and development in the United States. [Bloomberg]

🦾 Emerging Tech

Circle files for IPO as crypto market shows renewed investor interest - Stablecoin issuer Circle has confidentially filed for a public offering, signaling growing confidence in crypto amid improving market conditions. [CNBC]

Meta’s $1,000+ smart glasses will feature screen and AI assistant - Meta’s next-generation smart glasses, expected later this year, will include a visual display and AI capabilities in a sleek Ray-Ban frame. [Bloomberg]

Lightmatter unveils optical tech to supercharge AI chip communication - Lightmatter debuts new photonic interconnect and chiplet technologies designed to vastly improve bandwidth and energy efficiency in AI hardware. [Reuters]

🔬 Research

MoCha model enables realistic talking characters from audio input - Researchers introduce MoCha, an end-to-end model that generates expressive, movie-quality talking head videos from speech and text inputs. [arXiv]

⚖ Legal

Trump’s AI Action Plan draws nearly 9,000 public comments - The public weighs in on the forthcoming national AI policy, with responses ranging from tech firms to artists calling for transparency and rights protections. [Emerging Tech Brew]

Trademark battle brews over Elon Musk’s chatbot name ‘Grok’ - Musk’s AI company xAI faces legal pressure from a startup claiming prior use of the “Grok” brand in an escalating trademark conflict. [Wired]

OpenAI allegedly trained models on paywalled O’Reilly Media content - Researchers assert that OpenAI’s models, including GPT-4o, may have used copyrighted, subscription-only O’Reilly books for training without permission. [TechCrunch]

🎱 Random

Tinder’s new ‘Game Game’ focuses on IRL connection comfort - Tinder introduces a pre-date game feature aimed at helping users ease into real-life meetings by playing quirky, low-stakes challenges. [Tinder Pressroom]

🔌 Plug-Into-This

Hollywood Studios Profit from AI-Generated Fake Movie Trailers on YouTube

Major studios including Warner Bros. Discovery, Paramount, and Sony have been quietly collecting ad revenue from viral AI-generated fake movie trailers on YouTube, rather than removing them. This monetization tactic has sparked criticism from SAG-AFTRA, as it bypasses actors’ consent while exploiting their likenesses.

• YouTube channels like Screen Culture used generative AI to fabricate trailers for non-existent films, using real actors’ appearances and footage from unrelated projects.

• Instead of issuing copyright takedowns, studios opted to monetize these videos, claiming the ad revenue through YouTube’s Content ID system.

• YouTube has now removed these channels from its Partner Program, citing violations around misleading or reused content.

• SAG-AFTRA condemned the studios’ actions, arguing that they normalize the commercial exploitation of actors’ images without approval.

• The spread of tools like OpenAI’s Sora and Google Veo has made it easier for non-professionals to flood the platform with convincing fake media.

🤔 The studios’ choice to monetize rather than moderate AI-generated fan content could reveal a meaningful shift in strategy—rather than fight the AI boom, simply prioritize digital revenue and online engagement vs. enforcing traditional IP and performer protections.

Runway’s Gen-4 AI Solves the Character Consistency Challenge, Making AI Filmmaking Actually Useful

Runway has introduced Gen-4, its newest AI video model that brings character and scene consistency across multiple shots—addressing a key barrier to narrative coherence in AI filmmaking. This leap positions Runway as a leading player in generative video tools for storytelling.

• Gen-4 uses memory mechanisms to maintain identifiable character traits and environmental elements across different angles and frames.

• Current outputs are five to ten seconds at 720p, but the company is actively developing higher resolutions and longer durations.

• The release builds on prior Runway models like Gen-3 Alpha Turbo and Act-One, which handled advanced movement and face animation.

• It allows creators to build more continuous, professional video sequences, appealing to filmmakers and content studios exploring AI pipelines.

• The update aligns with a broader trend toward multi-shot consistency, following OpenAI’s recent visual character coherence upgrades.

🎥 Gen-4 represents a major shift from one-shot AI clips to sequence-based storytelling (without needing to build your own custom workflow). Generative video is evolving from novelty to viable production tool—especially for short-form and experimental formats.

ChatGPT Isn’t the Only Chatbot That’s Gaining Users

While OpenAI’s ChatGPT retains its lead with over 500 million weekly users, rivals like Google’s Gemini, Microsoft’s Copilot, Anthropic’s Claude, and xAI’s Grok are gaining traction. The chatbot market is quickly diversifying, driven by new features and aggressive platform integration.

• Gemini Daily visits rose 7.4% in March to 10.9 million, reflecting growing engagement post-model updates.

• Microsoft Copilot logged 2.4 million daily visits, aided by Office integration and workflow use cases.

• Claude averaged 3.3 million daily visits, while xAI’s Grok matched DeepSeek with 16.5 million, partly via X platform embedding.

• Most growth is tied to recent model rollouts, new UX features, and multimodal capabilities across ecosystems.

• Despite the competition, ChatGPT still commands roughly 10x more weekly app users than Gemini and Claude combined.

📈 Despite ChatGPT’s preeminence, the chatbot race is now looking to be less about singular dominance and more about ecosystem lock-in, where cross-platform utility, niche use cases, and native integration will determine long-term user loyalty across a fragmented AI landscape.

 🆕 Updates

📽️ Daily Demo

🗣️ Discourse