In partnership with

You spend two weeks on a report, send it to the team, and maybe one person reads past the summary. 

The problem is that nobody has 40 minutes to give you, but almost everybody has 60 seconds.

Google just shipped the fix:

With Gemini Notebook, you can now turn any dense document into a narrated 60-second video.

Upload the report, click twice, and out comes a short clip with visuals, narration, and the points that actually matter. Built entirely from your document.

Today I'm walking you through it, upload to share 👇️

Win AI Search Without a Big Team

92% of VCs use AI to find companies. 58% of buyers start there, too. If you're not showing up in AI answers, you're invisible before the conversation even starts. Join HubSpot for Startups, Anthropic, and Marketing Against the Grain on July 30 (11 am ET) for a live AEO teardown. Real startup. Real recs. Register and unlock the free Startup Visibility Bundle.

Register Now

What Gemini Notebook is (and what just shipped)

Quick catch-up if you've never touched it: Gemini Notebook is Google's free research tool. 

You upload your own documents into a "notebook," and the AI works only from what you gave it. It summarizes and explains your material instead of improvising from the open internet.

In early July, Google added Short Video Overviews: 60-second vertical videos (phone-shaped, like a Reel or a TikTok) generated straight from your sources. AI-made visuals, animations, and a narrator walking through the key points.

Under the hood it runs on Nano Banana 2 Lite, Google's fast image model, which is why the visuals generate in minutes instead of hours. 

The Short format launched with Google AI Pro and Ultra subscribers and is now reaching free accounts too. 

It's a gradual rollout, so if you don't see it yet, check back in a few days, or use the longer Video Overview format in the meantime. The steps below work the same way. 

Make your first one in five minutes

Grab a real document. The densest, most skimmed-over thing you've sent this month is the perfect test.

  1. Go to notebooklm.google.com and sign in with your Google account. Create a new notebook.

  1. Upload your document. PDF, Google Doc, Word file, or even a link to a webpage. One document is ideal for a first run: focused input, focused video.

  1. Open the Studio panel (the right-hand side of the screen) and choose Video Overview.

  1. Set the format to Short. You can also pick a visual style (Whiteboard, Watercolor, Anime, or classic) or let it choose for you.

  1. Steer it before you generate. Click the edit option on the video card and tell it who the video is for and what to emphasize. Something like: "This is for the sales team. Focus on the pricing change and what it means for renewals. Skip the methodology."

  1. Hit Generate. It builds in the background over a few minutes, so go do something else. Come back, watch it, download it, drop it in Slack or the team channel.

That's the whole workflow. Document in, watchable video out, five minutes of your attention total.

Before you hit share…

Three guardrails worth keeping:

  • Watch it against the source. The video is a compression, and compression means choices.

  • Check your company's rules before uploading confidential documents to this or any AI tool. Same rule as always, worth repeating.

  • Short videos are English-only at launch. If your team works in another language, the longer Video Overview format supports 80+ languages.

Use This Prompt Formula

Paste this into the video's customization box before you generate, and fill in the blanks 👇️


I'm turning this document into a 60-second video for [who will watch it, e.g., "my sales team"].

Focus on [the two or three points that matter most, e.g., "the Q3 pricing change and what it means for renewals"].

Skip background, history, and methodology. Use plain language that can be understood on the first listen. Keep the tone [professional / upbeat / urgent]. End with the single thing viewers should do next: [the action, e.g., "review their renewal list by Friday"].

Regenerating is cheap and fast, so treat the first version as a draft.

Tweak the instructions, run it again, and stop when you’re satisfied with the result.

Introducing The First Agentic CRM

Get revenue agents, workflows, and automations across every stage of your motion. Access customer data in real time through Attio's web app, MCP, API, and SDK.

Then Ask Attio anything about your business and get instant answers.

It's the CRM that runs the work behind every win.

Try Attio Now

Your roundup of the latest model releases and updates from the biggest AI labs.

  1. Google refreshed its everyday models and teased Gemini 4. On July 21, Google released Gemini 3.6 Flash and 3.5 Flash-Lite, the fast, cheap "workhorse" models that quietly power most of its AI features (including tools like Gemini Notebook). The new Flash is noticeably better at multi-step tasks and cheaper for the companies building on it, which usually shows up for you as snappier, more capable tools. Google also confirmed the bigger picture: flagship Gemini 3.5 Pro still isn't ready, and training has already begun on Gemini 4. (9to5Google) 

  1. Mira Murati's startup just shipped its first model. Thinking Machines, the lab founded by OpenAI's former tech chief, released Inkling on July 15. It's an "open-weight" model (one companies can download and run on their own computers instead of renting access) and it debuted as the top US open model on independent testing. The company's bet is the opposite of one giant assistant for everyone: businesses taking a strong base model and tuning it into their own. One of the most-watched startups in AI finally showed its cards. (TechCrunch) 

  1. Apple is suing OpenAI over its secret hardware project. Filed July 10, the lawsuit alleges former Apple executives now at OpenAI took confidential files and supplier information on their way out. OpenAI has hired 400+ former Apple employees and bought Jony Ive's design startup for $6.5B as it builds a consumer AI device. Why it matters: the next fight in AI is over the gadget in your pocket, and two of the biggest players just went to court over it. (TechCrunch)

Advertise with Build with AI

Get in front of an audience of professionals using AI day-to-day: founders, engineers, operators, and product builders.

Interested in advertising? Respond to this email for rates and details.

Until next time,

William Ryan

Editor-in-Chief @ Build with AI

PS: Follow me on X for daily updates and AI workflows.

Keep Reading