NotebookLM’s Video Overviews is now available in your language

After ​launching​ last month, Video Overviews just received a major upgrade, now supporting over ​80 languages​. Users can turn notes, reports, or documents into narrated slide-deck videos, and learn in the language they’re most comfortable with.

The latest update adds a few upgrades worth knowing:

• You can now generate narrated videos in 80+ languages, including Vietnamese, Arabic, Hindi, and Swahili. Helpful if you’re working across regions or training local teams.
• It now handles longer and more complex documents, including full reports, PDFs, or product guides.
• You can create multiple outputs from the same notebook: video, audio, mind map, written summary.
• Studio lets you set the video’s structure, tone, and audience so you’re not just getting a random summary, but something tailored to how you want to communicate.
• And you can listen to an audio overview while reviewing other views, like the mind map.

If you’re new to NotebookLM: it’s Google’s AI-powered workspace for working with your own documents, kind of like having an AI research assistant that knows your files. You drop in notes, slides, PDFs, transcripts, or links, and it can answer questions, summarize, generate talking points, audio overview and now… video overview with narration.

This week, we tested it out for something practical: Turn a document into a narrated how-to video your team can actually use.

Here's how we can translate ChatGPT's guide on effectively prompting ChatGPT-5 into a concise and engaging how-to video:

1. Start a new project with the content you want to train: Paste in this link as your first source ​https://cookbook.openai.com/examples/gpt-5/gpt-5_prompting_guide​.
2. Customize the video’s structure and language: Tell NotebookLM exactly how you’d like it to present the information, your video's objectives, and the language you want it to be in.
3. Generate the Video Overview: Once you’re happy with the setup, it’s just one click to start generating. You can review, adjust the customization, and then regenerate if needed.
4. Download and share with your team.
5. Rethink what else you could turn into a training video: An internal SOP into a quick refresher, product guides into client-facing explainer video, onboarding videos in multiple languages.

Need a bit more detail? Check the ​full tutorial here​.

This is just one example of how AI can elevate your content, starting with a single spark. To go further and build systems around this, explore our ​​Build Your AI Team​​ course for practical guidance.

Your AI Team: Claude in Chrome’s extension, Google Vids’ AI enhancement, Gemini’s image editing, and more.

Every week, I report on the top updates to your favorite AI tools. This week:

You can bring Claude into any webpage

Anthropic just launched a new Chrome extension that makes Claude available wherever you browse, turning the chatbot into something more like a live research assistant.

What makes it special:

Claude is no longer limited to its own interface. With the new Chrome extension, you can now open Claude in a sidebar on any website, including Google Docs, PDFs, Notion, and more.

It’s context-aware, too. If you highlight text on a page, Claude will automatically pick it up as context for the conversation. That means you can do things like:

• Summarize a complex paper without pasting it in
• Draft emails for you
• Ask Claude to rewrite or explain a paragraph you’re working on
• Chat through ideas while browsing research, articles, or slides

It’s currently in controlled preview, limited to just 1,000 Max plan subscribers. If you’re curious to try it, you can​ join the waitlist​.

Quick Hits from your favorite AI tools:

• Google Vids ​goes public​ with AI enhancements: Google’s AI-powered video editor Vids is no longer limited to Workspace users; it’s now available broadly and includes tools like storyboard generation, 12 prebuilt avatars, 8-second video creation from images, and auto-removal of filler words and pauses.
• WhatsApp introduces private ​tone-tuning​ for messages: The new “Writing Help” feature uses Meta’s Private Processing to suggest tone-adjusted rewrites in styles like professional, funny, or supportive right in the chat, and processed entirely on-device for full privacy.
• Gemini’s image editing ​goes bananas​ (in a good way): Google’s Gemini app now includes a DeepMind-powered “Gemini 2.5 Flash Image” model, nicknamed “nano‑banana,” which lets you seamlessly change backgrounds, combine images, or tweak styles, all while preserving subject likeness and stamping outputs with visible and invisible watermarks.

Read more news at the end.

AI Tools Recommendation: AI for Image Editing and Generation

Instead of chasing the latest hype, I’m using this section to highlight the top AI tools in key categories, only those Lead with AI community members are actually using and relying on.

This week: AI for Image Editing and Generation

Need to create visuals without opening Photoshop? These free, beginner-friendly tools enable you to generate and edit images using simple text prompts, making them ideal for pitch decks, social media, or prototypes.

​Canva AI​

Canva’s AI features are built into a platform millions already use.

• Magic Edit lets you remove or replace objects by describing what you want
• Text to Image can generate graphics based on your prompt
• Magic Design creates entire visuals or presentations from a few inputs

​Leonardo.Ai​

Leonardo is one of the most popular platforms for high-quality image generation, especially for product renders, concept art, and digital design.

• Choose from different model styles (photo-realistic, 3D, anime, etc.)
• Fine-tune outputs with control tools like prompt weighting and image upscaling
• Used by creators, game designers, and small businesses alike

​4o Image Generation​

OpenAI’s latest image model is built directly into GPT‑4o, making it faster, more accurate, and easier to edit, right inside the ChatGPT experience.

• Generate high-quality images instantly from a simple prompt
• Click on any part of the image to edit it with natural language instructions
• Improved visual fidelity and alignment, with faster rendering and built-in safety filters

Gemini and Grok are catching up to ChatGPT, what to do when AI investments haven’t paid off, along with more crucial AI stories

Every day, Daan, Wendy, and I read all the AI news so that you don't have to.

Here are the must-read stories of the week:

​Google and Grok are catching up to ChatGPT, says a16z’s latest AI report​

Andreessen Horowitz’s new benchmark study shows Gemini and Grok narrowing the gap with ChatGPT across reasoning, coding, and math, especially in real-time and multilingual tasks.

​My AI Investments Haven’t Paid Off. Should I Tap the Brakes?​

This HBR piece explores how leaders can recalibrate AI strategies when early returns underdeliver, without losing sight of long-term value.

​Chatbots can be manipulated through flattery and peer pressure​

A new study finds that even top AI models like GPT‑4 can change their responses based on emotional cues, raising questions about how easily they can be manipulated.