🧠 Thinking Machines co-founder joins Meta

Good morning, AI enthusiast. Another big move in Silicon Valley’s AI talent wars, Meta just poached one of Mira Murati’s top co-founders.

In today’s AI newsletter:

• Meta hires Thinking Machines Lab co-founder Andrew Tulloch

• xAI is building an AI-generated game by 2026

• Sam Altman questions what “real work” even means

• How to automate code review and fix issues before commit

• 4 new AI tools & more

AI TALENT

🧠 Meta Hires Thinking Machines Lab Co-founder

Andrew Tulloch, co-founder of Thinking Machines Lab (the AI startup led by former OpenAI CTO Mira Murati), is heading to Meta, according to The Wall Street Journal.

→ Tulloch confirmed his departure in a message to employees
→ Meta reportedly offered him a package worth up to $1.5B over six years
→ Follows Meta’s failed attempt to acquire Thinking Machines Lab earlier this year
→ Tulloch previously worked at OpenAI and Meta’s AI Research Group (FAIR)

Why is important: This marks another high-profile win for Meta as Zuckerberg doubles down on AI recruiting. The move highlights how fiercely tech giants are competing for elite AI talent, even if it means billion-dollar offers.

AI NEWS

🤖 xAI Is Building an AI-Generated Game by 2026

Elon Musk’s xAI is developing “world models” capable of generating interactive 3D gaming environments, aiming to launch a playable AI-created game before 2026.

→ Nvidia veterans Zeeshan Patel and Ethan He joined xAI to lead physics-based world model development
→ The company is hiring for its new “omni team” and a “video games tutor” to train Grok on game design
→ Musk said the goal is to release a “great AI-generated game” by the end of next year, with AAA-level quality
→ World models like Genie 3 have already shown promise, but intuitive control and logic remain the next frontier

Why is important: Musk wants xAI to merge gaming, physics, and generative AI, not just to make a game, but to build a foundation for AI that can understand and simulate the real world.

AI TALKS

⚡ Sam Altman Questions What “Real Work” Even Means

At OpenAI’s DevDay, CEO Sam Altman suggested that if AI wipes out jobs, maybe they weren’t “real work” to begin with. Drawing on a farmer analogy, he said a farmer from 50 years ago would see most modern jobs, like coding or marketing, as “playing a game to fill your time.”

→ Altman says AI might make us rethink what qualifies as meaningful work
→ Compares today’s knowledge jobs to the “fake work” of the future
→ Believes human creativity will always find new outlets
→ Adds he’s “less worried” about job loss, “more worried in other ways”

Why is important: Altman’s comments highlight how AI is reshaping the definition of work itself. But they also reveal a growing divide, between tech optimism and the real-world anxiety of millions facing automation.

HOW TO AI

🎨 How to Automate Code Review and Fix Issues Before Commit

CodeRabbit is an AI code reviewer that not only writes code but also reviews, debugs, and secures it before you commit. With its new MCP integration, it can connect context across your entire development ecosystem, making code safer and more reliable.

🧰 Who is This For

• Developers using AI coding assistants like Cursor, Claude, or Copilot

• Teams looking to ensure production-ready commits

• Startups and engineering teams wanting automated security and logic checks

• Engineers who want to prevent costly bugs before deployment

STEP 1: Install the CodeRabbit

Open your IDE, VS Code, Cursor, or Windsurf. For this tutorial will use Cursor.
In the left-hand activity bar, click the Extensions icon. In the search bar at the top of the panel, type CodeRabbit, then press Enter.

When the CodeRabbit extension appears, click on its name to open the details page, and then click Install in the upper-right corner. Once the installation is complete, a CodeRabbit tab or icon will appear in your sidebar.

STEP 2: Link Your Account

After installation, open the CodeRabbit tab in your IDE. You’ll see an option to Sign In or Link Account. Click this, and your browser will open automatically to the CodeRabbit login page.

Sign in using your Git provider (for example, GitHub or GitLab). Review the permissions requested and confirm authorization. Once authenticated, return to your IDE, the CodeRabbit tab will now display your connected account status.

STEP 3: Review and Apply Fixes

With your project open, navigate to the CodeRabbit tab and click Review. The system will analyze your code in real time, scanning for logic flaws, syntax issues, and potential vulnerabilities.

Once the review is complete, you’ll see a list of suggestions organized by severity. Click any item to jump directly to the affected line of code. Each suggestion includes a short explanation and, where possible, a one-click fix option.

Click Apply Fix to automatically update your code based on the recommendation. You can re-run the review at any time to validate the new version.

STEP 4: Enable MCP for Context-Aware Reviews

To activate the new MCP integration, open the CodeRabbit Dashboard in your browser. From the sidebar, select Integrations, then click the MCP Server tab.

Click New MCP Server to add a connection. Choose your external tool—such as Jira, Linear, or Notion, and enter the server URL and authentication credentials. Once connected, CodeRabbit will automatically sync task and issue context.

When creating a pull request on GitHub or GitLab, include the corresponding issue ID from your connected tool (for example, “Fixes PROJ-142”). CodeRabbit will use this data to enhance its analysis and deliver context-aware feedback directly in your pull request comments.

Review the CodeRabbit comments, address the flagged issues locally, and push the updated commits. CodeRabbit will re-review automatically, ensuring your code is both secure and aligned with your project requirements. Once satisfied, complete the merge into your main branch.

ESSENTIAL BITES

• GPT-5 and Gemini 2.5 earned gold medals at the International Olympiad on Astronomy & Astrophysics, achieving top performance among several models.

• OpenAI is under legal scrutiny as authors and publishers access internal records over its use of pirated books from Library Genesis.

• OpenAI reports GPT-5 is 30% less politically biased than earlier models, based on 500 test prompts.

• Apple is close to acquiring computer vision startup Prompt AI, adding its 11-person team and tech to Apple’s smart home division.

HOT AI TOOLS

⚙️ Claude Code: Anthropic’s AI coding assistant

📱 Apps SDK: Build & chat with apps inside ChatGPT

💼 Gemini Enterprise: Create & run AI agents

✨ Firstbuild: Create custom apps with open-source AI

THAT’S IT FOR TODAY

Thanks for making it to the end! I put my heart into every email I send, I hope you are enjoying it. If you are, you can support by sending it to your friends, so we all grow together! Let me know your thoughts so I can make the next one even better.

See you tomorrow :)

- Dr. Alvaro Cintas