Today’s menu 🍽️👇️ 

Length: 5 minutes ⏰ 

• NotebookLM turns research into video 🎬

• Nvidia pulls back from AI partners 🤔

• Google Search gets Gemini Canvas⚡

• Claude Code gets voice mode 🗣️

• Midjourney for killer thumbnails 🖼️

• Prompt of the week 💬

• Top AI tools of the week ⚒️

FRESH AI NEWS

NotebookLM now creates video briefs 📚

Google upgraded NotebookLM with a feature that converts research notes and sources into cinematic AI video summaries, turning documents into visual explainers automatically 🎬

• NotebookLM can now generate short documentary-style videos that explain research topics using visuals, narration, and structure 🎥

• The AI reads uploaded PDFs, notes, and links, then extracts key insights and turns them into a visual storyline 📄

• Videos include AI voice narration and scene transitions, making complex topics easier to understand 🧠

• The tool automatically adds charts, diagrams, and contextual imagery to illustrate important concepts 🖼️

• The feature turns NotebookLM from a study assistant into a content-creation tool for presentations, teaching, and explainers 🌐

Why You Should Care: AI research tools are advancing from text summarization to creating multimedia explanations. NotebookLM transforms documents into cinematic videos, making complex information more accessible and shareable, potentially changing how students, analysts, and creators present research.

Nvidia steps back from OpenAI 🚨 

Nvidia CEO Jensen Huang suggested the company is scaling back its involvement with OpenAI and Anthropic, but his vague explanation has sparked confusion about what the shift actually means 🤔

• Huang said Nvidia is reducing direct engagement with OpenAI and Anthropic, though he didn’t clarify whether this means fewer investments, partnerships, or technical collaborations 🗣️

• Despite the remarks, Nvidia remains the primary GPU provider for many leading AI labs, including both companies 💻

• Huang suggested Nvidia wants to avoid being overly tied to specific model developers as the AI ecosystem expands 📊

• Nvidia is increasingly supporting thousands of AI startups, enterprises, and sovereign AI initiatives, rather than concentrating on a few labs 🌍

• Analysts noted Huang’s comments lacked detail, leaving uncertainty about whether this is a real strategic shift or simply positioning 🤷

• Investors and industry watchers are trying to interpret whether Nvidia is diversifying influence or quietly stepping back from key partners 📈

Why You Should Care: Nvidia is central to the AI economy as most major models use its GPUs. If it reduces reliance on a few leading labs, it could shift power towards a more distributed AI ecosystem, with Nvidia supporting numerous AI companies instead of just a few dominant ones.

Google adds Gemini Canvas to AI Mode 🤖

Google has expanded Gemini’s Canvas workspace inside Search AI Mode to all U.S. users, turning Google Search into a place where you can create documents, tools, and projects directly from prompts 🧠

• Google has rolled out Gemini Canvas in AI Mode to all users in the U.S. (English), moving it beyond its earlier Google Labs testing phase 🧠

• Canvas lets users draft documents, build simple apps, plan projects, and generate code directly within the search interface 📝

• When activated, Canvas opens a side workspace where AI outputs can become interactive dashboards, prototypes, or structured documents 🧰

• Users can refine content through prompts, adjusting structure, tone, or logic without leaving Google Search 🔄

• The tool can generate emails, blog posts, reports, and even HTML/JavaScript tools that run in the browser 🧑‍💻

• The update represents Google’s push to make Search a productivity hub, not just a place to find links 🚀

Why You Should Care: Google is transforming Search into a creation engine by integrating Canvas in AI Mode, enabling users to move from research to drafting and tool-building within the app. This positions Google Search as a competitor to AI workspaces like ChatGPT and Claude Artifacts, suggesting that the future of search involves doing work, not just finding answers.

Voice mode arrives in Claude Code ✨ 

Anthropic has added voice mode to Claude Code, letting developers talk to the AI while coding, turning the coding assistant into a conversational pair-programming partner 🎙️

• Developers can now speak directly to Claude Code instead of typing prompts when working on projects 🎙️

• Users can ask Claude to explain code, debug errors, generate functions, or refactor files using voice commands 💻

• Voice interaction enables quicker iteration when brainstorming architecture or troubleshooting issues ⚡

• Claude still retains project context and codebase understanding, so spoken instructions apply to the active workspace 🧠

• The feature is designed for continuous coding sessions, where developers can keep working while giving verbal instructions 🔄

• Voice mode complements Claude Code’s broader goal of becoming a full developer agent, not just a chat interface 🚀

Why You Should Care: Coding assistants are moving toward real-time collaboration rather than static prompts. Voice interaction makes AI feel more like a human teammate during development, potentially speeding up ideation and debugging while making AI-assisted programming more natural for long coding sessions.

AI THUMBNAIL WORKSHOP

Create viral thumbnails with MJ 🚀

Learn workflow for creating eye-catching thumbnails that look intentional and professional, not like generic AI art. By combining smart prompts, quick image iteration, and simple design tweaks, you can produce standout visuals for videos, blogs, or social posts in minutes ✅

Here’s how to get started 👇️ 

1. Open Claude or another chat tool, describe your content topic, and ask for several image prompts formatted like this:

   [Subject + facial expression] + [object or action] + [environment] + [lighting style] + [framing with room for text]. Aspect ratio 16:9.

2. Paste each prompt into the Midjourney web interface. When you see a concept that works, use the Vary options (subtle or strong) to refine it until the composition feels balanced.

3. Import the image into a design tool like Canva and keep the overlay minimal, use only a few powerful words that clearly communicate the hook.

4. Choose a bold, easy-to-read font, slightly reduce opacity so the text blends with the image, and stick to high-contrast colors like white or yellow. Adding a thin outline can help the words pop against busy backgrounds.

Pro Tip: If your image doesn’t leave enough space for text, ask your AI assistant to regenerate the prompt with intentional negative space in the composition so the headline fits naturally without covering the main subject ✨

The best place on the Internet to learn about AI! 🌐 🤖 

I just started my own X page where I provide AI news, guides, hacks and more (for FREE)!

We don’t miss a day, that means you get the most recent AI news, practical workflows, expert guides and premium content everyday.

TOOLS OF THE WEEK

These AI tools are going VIRAL 🐝 

Viktor - OpenClaw for enterprise with 3,000+ integrations in Slack and Microsoft Teams, delivering board-ready PDFs, dashboards, and web apps 🤖 

GPT-5.3 Instant - OAI’s ChatGPT default model update with fewer refusals and less hallucinations 🚀 

Codex - OAI’s app interface for managing coding agents, now on Windows 🧮 

Grok Imagine - xAI's image and video generation platform 🌌