We’re Not Just a Newsletter Anymore. We’re an AI Education Platform.
Digestible AI is now Digestibly, because we’re not just helping you read about AI anymore. We’re helping you work smarter with it.
For the past year, we’ve published the newsletter thousands of knowledge workers rely on to stay ahead in the age of AI. But across industries, we kept seeing the same thing: smart, capable people getting left behind, not because they weren’t good at their jobs, but because they hadn’t been trained on the tools reshaping them.
While enterprise teams brought in consultants and quietly upskilled their operators, the average professional was told to "figure it out" or risk falling behind. And too often, they were replaced by someone who knew how to use AI.
That’s the gap we’re here to close.
Our mission?
“To shrink the knowledge divide by teaching one million people AI.”
That starts with the launch of our University, our always-evolving education platform built for professionals, creatives, founders, and anyone using their mind to make a living.
You told us what a fair price would look like:
75+ lessons covering AI fundamentals, agent basics, LLMs, prompts, and automations
Weekly updates to all existing lessons as AI evolves
Brand-new module (20+ lessons) released every month
2x Monthly live Q&A calls (listen in & get your questions answered)
Over half of you said you’d pay $29/month for that.
So we’re launching at just $19/month for early supporters.
Full access opens August 23rd.
We’re Digestibly. And we’re just getting started.
Now, back to your regularly scheduled newsletter.
Google's Genie 3 creates interactive 3D worlds in real-time, OpenAI releases their first open-source models since GPT-2, and NotebookLM adds video overview capabilities...
In this edition we’ll be covering…
Google DeepMind's revolutionary Genie 3 world model
OpenAI's surprise release of GPT-OSS open-source reasoning models
A tutorial on creating video overviews with NotebookLM's new Studio features
5 trending AI signals from government approvals to safety concerns
3 cutting-edge tools including Claude's latest upgrade
And much more…
The Latest in AI
Google’s Genie 3 Is Building Reality From Scratch
Image from: Google
Google DeepMind just announced Genie 3, and honestly, it's hard to wrap your head around what this thing can actually do…
The AI can generate interactive 3D worlds in real-time at 24 frames per second that you can actually navigate and explore for several minutes straight!
What's coming with Genie 3?
Real-time world generation, so you can type a prompt and get a fully navigable environment that responds to your actions instantly at 720p resolution.
Unlike previous models, Genie 3 maintains visual memory for up to a minute, so when you revisit locations, they look exactly the same as before.
You can literally change the world on command - alter weather conditions, introduce new objects, or modify the environment with text prompts mid-experience.
The company tested their SIMA agent in these generated worlds, proving these environments can actually train AI agents for complex tasks.
So What?
Genie 3 represents a massive leap from simple video generation to interactive world simulation. While we're still in the early days, this could fundamentally change how we think about virtual environments, training AI agents, and even education.
Google's betting that world models are a key stepping stone to AGI, and frankly, after seeing what Genie 3 can do, they might be onto something.
The implications go way beyond gaming - imagine training autonomous vehicles in generated scenarios, or teaching AI systems complex real-world tasks in simulated environments that feel completely authentic.
The model is still in research preview mode with limited access to academics and creators, so don't expect to be building virtual worlds from your laptop just yet!
Industry Intel
How to Run OpenAI’s New Open-Source Models with Ollama
Plot twist of the year. OpenAI just released their first open-source language models since GPT-2.
Meet GPT-OSS-120B and GPT-OSS-20B - two reasoning models you can now run locally on your own hardware.
Here's how to get these models running locally using Ollama:
Install Ollama if you don't have it already by visiting ollama.com and downloading it for your operating system.
Pull the models you want:
ollama pull gpt-oss:20b
ollama pull gpt-oss:120bStart chatting locally:
ollama run gpt-oss:20b🔥 Power User Tip: The 20B model runs well on most desktops with 16GB of memory, while the 120B model needs about 80GB of memory but delivers performance comparable to OpenAI's o4-mini model. Both models support chain-of-thought reasoning and tool use, making them incredibly versatile for local development.
Everything AI, All in One Place
Pre-sell ends Aug 23—then price jumps from $19/mo to $29/mo (35% increase!). Lock in:
75+ lessons on AI fundamentals, agents, LLMs, prompts & automations
Weekly updates as AI evolves
20+ new lessons released each month
2× monthly live Q&A calls (get your questions answered in real time)
Cancel anytime
Get Your Hands Dirty!
Create Video Overviews in NotebookLM
Google just dropped a fire update to NotebookLM that transforms your documents into narrated video presentations.
Think of it as turning your research into a personal documentary with slides.
Here's how to make video overviews:
Upload your content to NotebookLM - this can be PDFs, documents, websites, or any other supported format.
Navigate to the Studio panel which now has a fresh new design with four distinct tiles at the top.
Click on "Video Overview" and let the AI analyze your content to create narrated slides with visuals, diagrams, and key quotes from your documents.
Customize your overview by specifying topics to focus on, your learning goals, or describing your target audience in the prompt box.
Quick Bites
Stay updated with our favorite highlights, dive in for a full flavor of the coverage!
Google announces free AI Pro access for college students providing 12 months of Gemini 2.5 Pro, NotebookLM, and other premium features to students 18+ in the US, Japan, Indonesia, Korea, and Brazil, with plans to expand globally.
Wells Fargo deploys AI agents business-wide through partnership with Google Cloud's AgentSpace, marking one of the largest enterprise rollouts of agentic AI across banking operations from branch tellers to investment banking teams.
US government approves OpenAI, Google, and Anthropic for federal AI vendor list with ChatGPT, Gemini, and Claude now officially approved for use by civilian federal agencies as the Trump administration pushes AI adoption across government.
BBC reports on AI safety concerns as researchers warn that rapid AI advancement could outpace safety measures, with calls for more robust testing and evaluation frameworks before deploying increasingly powerful AI systems.
One of the many fears about AI use becoming widespread is that people can now alter images – sometimes convincingly – without any technical skills. An example of this surfaced recently when an Airbnb guest said a host manipulated photos in a false £12,000 ($9,041) damage claim.
Trending Tools
🤖 Claude 4.1 Opus - Anthropic's latest flagship model with major upgrades to agentic tasks, real-world coding, and reasoning capabilities, now available to paid Claude users and through the API.
🎙️ Transcript LOL - AI-powered transcription tool that converts audio and video files into text with 99.8% accuracy, supporting speaker recognition and generating summaries from podcasts, meetings, and videos.
⚡ Trickle - No-code platform that turns ideas into live apps and websites using AI, featuring built-in database functionality, calendar sync, and task management capabilities for rapid prototyping.
Let us know!
Are you signing up for our brand new AI University?
Until we Type Again…
Thank you for reading yet another edition!
How did we do?
If you have any suggestions or specific feedback, simply reply to this email or fill out this form. Additionally, if you found this insightful, don't hesitate to engage with us on our socials and forward this over to your friends!