Grok Imagine, Project Genie & AI in Hollywood

Morning Brief: xAI, Google DeepMind, Claude Workflows, and AI in Hollywood

Good morning, {{ first_name | AI enthusiasts }}. Today's AI roundup covers major moves in AI video, interactive world generation, competitor intelligence automation, and how filmmakers are adopting generative tools. This edition also highlights several new tools and practical workflows relevant to creators, developers, and anyone following gaming news and AI adoption in media.

xAI: Grok Imagine Tops Video Leaderboards

The news: Elon Musk's xAI launched the Grok Imagine API, a text-to-video, image-to-video, and video-editing suite that rapidly climbed to No. 1 on key leaderboards for video.

Key details

• Features: Text-to-video, image-to-video, and editing with up to 15-second clips and native audio.
• Pricing: $4.20 per minute (audio included). That undercuts competitors like Veo 3.1 ($12/min) and Sora 2 Pro ($30/min).
• Editing tools: Swap objects, restyle scenes, animate characters with custom performances, and change environments on command.
• Rankings: Debuted at No. 1 on Artificial Analysis' text and image-to-video lists; trails Veo 3 and Sora Pro in Arena's Video Arena.

Why it matters: Grok Imagine's aggressive pricing plus leaderboard performance could make it the go-to option for creators and developers who need fast, affordable iterations.

Example use cases: indie game studios creating short cinematic cutscenes, social media creators producing quick story-driven clips, and app developers integrating automated video previews.

Google DeepMind: Project Genie Goes Public

The news: Google DeepMind released Project Genie, a web app powered by the Genie 3 model that generates explorable, AI-built worlds in real time.

How it works

• Users prompt a setting and character, preview results via Nano Banana Pro and Gemini, then navigate in first- or third-person.
• Characters can walk, fly, or drive; the model remembers previously generated areas to keep visual consistency.
• Sessions currently cap at 60 seconds due to compute costs; each user gets a dedicated compute "chip" for the session.
• Rollout: initially available to Google AI Ultra tier subscribers ($250/mo), with broader access planned later.

Why it matters: Real-time world simulators open possibilities for robotics training, game prototyping, and architectural visualization. As fidelity rises, gaming news and simulation tech are converging.

AI Training: Build a Competitor Database with Claude Cowork

The news: A step-by-step guide shows how to create an automated competitor intelligence database using Claude Cowork. The system largely builds itself from downloadable prompt files.

Steps to set up

1. Download the free prompt files and save them in a new folder.
2. In the Claude desktop app (Mac only), open the Cowork tab and select the folder.
3. Ask Claude to read the SOP and set up your directory (2–3 minutes).
4. Test: provide a competitor name and website; Claude generates a "killsheet" and files it automatically.
5. Run a new killsheet each quarter or after major competitor events.

Pro tip: Ask Claude to create battlecards and comparison pages from each killsheet for quick team briefings.

AI & Hollywood: Darren Aronofsky's Revolutionary War Series

The news: Filmmaker Darren Aronofsky's Primordial Soup released "On This Day… 1776," a short-form series that reimagines events from the American Revolution using Google DeepMind visuals alongside SAG-AFTRA voice actors.

What to know

• Each episode drops on the 250th anniversary of the event it depicts and streams on TIME's YouTube channel.
• The project blends visuals with professional voice talent, emphasizing artist-led storytelling rather than full automation.
• Aronofsky previously collaborated with DeepMind on the Veo-assisted film ANCESTRA, shown at Tribeca.

Why it matters: AI video is moving into mainstream production workflows. From creative shorts to studio projects, the tech is shifting from novelty to a practical production tool.

Trending Tools and Integrations

• Speechmatics — voice tech with a Startup Program offering credits for production launches.
• Project Genie — DeepMind's interactive world generator for simulation and prototyping.
• Grok Imagine — xAI's video model and API for fast, low-cost video generation.
• Wonda — an AI agent for video editing and creative direction.

Other AI Headlines

• Microsoft announced Foundry, an Azure platform for building and governing AI apps and agents.
• Apple reportedly acquired Israeli AI audio startup Q AI for nearly $2B.
• Anthropic faces lawsuits from music companies claiming unauthorized use of songs to train Claude; plaintiffs seek $3B+.
• Nvidia, Microsoft, and Amazon may negotiate up to $60B in investments in OAI, which could value the company above $700B.

Community Workflow Spotlight

From Gabriel V. in Montreal: He built an automation with Microsoft Power Automate and the Claude API. Incoming emails are scanned; invoices get detected and uploaded to SharePoint with the right name. The result: dramatically less administrative overhead.

Closing

That's it for today's rundown. For creators, developers, and anyone tracking gaming news and AI-driven media, these updates show rapid progress in video generation, interactive worlds, and practical automations. Share how you're using AI — we may feature your workflow next.

See you soon,
Rowan, Joey, Zach, Shubham, and Jennifer — the humans behind The Rundown