Musk vs OpenAI Dismissed — Latest AI & Gaming News

Elon Musk vs. OpenAI: Jury Dismisses $100B Suit

Good morning, {{ first_name | AI enthusiasts }}. After three weeks of high-profile testimony, leaked messages, and more than $100 billion in claims, the blockbuster trial between Elon Musk and OpenAI ended without a dramatic verdict. A jury dismissed Musk's suit against OpenAI, Sam Altman, Greg Brockman, and Microsoft, ruling the case was filed too late.

Jury dismissed the case unanimously on statute-of-limitations grounds; Musk says he will appeal and calls the ruling a "calendar technicality."

What happened

The suit alleged Altman and Brockman essentially converted a nonprofit charity into a for-profit operation and that Microsoft unlawfully aided that shift. But the jury found Musk had known about the for-profit direction years earlier and waited too long to file. OpenAI's lawyers pointed out Musk had supported for-profit steps in the past and later founded his own rival, xAI, before suing in 2023.

Musk's claim against Microsoft was also dismissed. He argued Microsoft's multibillion-dollar backing helped OpenAI improperly profit, but the court again ruled the suit was untimely.

Why this matters

• Legal outcome: Big technical win for OpenAI, since the case was dismissed rather than decided on merits.
• Unresolved governance questions: The trial raised public debate about control, nonprofit vs. for-profit structures, and how billions affect mission-driven AI labs.
• Potential appeal: Musk has pledged to appeal, so a more definitive legal resolution could still come later.

Cursor's Composer 2.5: Efficient Coding Near the Frontier

Cursor released Composer 2.5, an in-house coding model based on Moonshot's Kimi K2.5. It reaches near-frontier benchmark scores while keeping token costs low.

Key details

• Performance: Composer 2.5 approaches Anthropic 4.7 and GPT-5.5 on development benchmarks and improves roughly 10% over Composer 2.0.
• Cost: Average CursorBench tasks cost under $1 with Composer 2.5, versus up to $11 per task for competing models at similar scores.
• Training: Composer 2.5 was partially trained on Colossus 2 compute; Cursor is training a larger SpaceXAI model with 10x more compute.

Why it matters: Teams that need strong developer-model performance at lower cost may switch to models like Composer 2.5. With Colossus compute behind Cursor, its next release could challenge current frontier models.

3D Modeling with Claude and Blender

Learn how to connect Blender to Claude Code using the MCP extension so you can create and edit 3D scenes with plain English prompts.

Quick setup

1. Install Blender and the MCP extension. Enable it under Edit > Preferences > Add-ons.
2. In your project folder, run: "claude mcp add blender -- uvx blender-mcp-server claude mcp get blender" to register Blender with Claude Code.
3. Ask Claude to verify the connection: "Connect Claude Code to Blender and make sure my mcp.json or Claude MCP config is set up correctly..."
4. Test: "Use Blender MCP to model my name in 3D. Add disco balls and lighting, reflective materials, and a camera angle like an event poster."

Pro tip: Download assets from BlenderKit, open them in Blender, and ask Claude to arrange objects, adjust lighting, and render. This workflow speeds up prototyping for game assets, product mockups, and virtual event visuals.

Odyssey's Multimodal, Multiplayer World Models

Odyssey unveiled two breakthroughs: Starchild-1, a real-time multimodal world model, and Agora-1, a multiplayer shared-world stream. Both hint at new possibilities for gaming news and interactive entertainment.

Highlights

• Starchild-1: Generates synchronized audio and video in real time and adapts continuously to user input.
• Agora-1: Hosts up to four players in one shared world with a persistent game state—demoed as a live GoldenEye-style simulation where every pixel is generated on the fly.

Why it matters: Moving from pre-rendered clips to live, adjustable shared streams opens new avenues for gaming, storytelling, and robotics simulation. For readers following gaming news, these systems could reshape multiplayer game design and AI-driven content generation.

Featured: You.com Guide — Beyond Latency When Choosing an API

Picking an API solely by latency benchmarks can mislead teams. This You.com guide explains why accuracy and real-world metrics matter more than raw p50 latency.

What you'll learn

• Why p50 latency hides real user failures.
• The "time-to-useful-result" framework for practical evaluation.
• Four hidden cost drivers that appear in logs, not vendor tables.
• How to evaluate APIs at real concurrency levels instead of demo conditions.

Trending AI Tools

• Moby 2 — AI ecommerce operator for Meta ads, Klaviyo campaigns, and creatives.
• Composer 2.5 — Cursor's efficient, near-frontier coding model.
• Krea 2 — Krea's first in-house image model, now generally available.
• Devin — New auto-triage security feature with long-term memory for code review.

Other Industry Moves

• Anthropic acquired Stainless, the team behind Claude's SDKs and MCP tooling.
• OpenAI partnered with Malta to offer ChatGPT Plus to citizens who complete a national AI literacy course.
• Amazon launched Alexa Podcasts, a NotebookLM-style custom podcast creator for Alexa+.
• Meta announced layoffs affecting up to 8,000 employees amid an AI efficiency push.
• OpenAI and Dell partnered to run Codex inside corporate data centers for enterprise coding workflows.

Community Workflow: Claude + Obsidian for Mindfulness Teaching

Reader Michael D. from Littleton, CO shares how he centralized years of notes into Obsidian and connected Claude to his vault. Claude pulls his real reflections to generate teaching material that stays grounded in lived experience rather than generic summaries. Michael reports new teaching perspectives and unexpected personal insights.

"Claude sometimes uses my notes in ways I wouldn't have thought of, giving me new perspectives on my own practice."

How to share your workflow

We feature reader workflows in every issue. Tell us how you use AI to work smarter or save time.

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