NVIDIA Goes Full Matrix

NVIDIA wants to link everything, agents are getting a UI glow-up, and GPT-4o just learned how to direct movies…

In this edition we’ll be covering…

• NVIDIA’s NVLink Fusion

• A breakdown on the new AG-UI protocol

• A tutorial on animating GPT-4o images with Kling

• 5 fast AI signals to keep you sharp

• 3 more AI tools to keep you productive

• And much more…

The Latest in AI

NVIDIA New Highway Lets More AI Chips Link Up

NVIDIA rolled into Computex 2025 and dropped NVLink Fusion, a new chip-to-chip superhighway for AI.

And for the first time ever, they’re letting other companies use it. It’s like NVIDIA handed out VIP passes to its most exclusive club…

What’s the Deal?

• NVLink Fusion allows companies like Marvell, MediaTek, Fujitsu, and Qualcomm to integrate their CPUs and AI chips with NVIDIA’s GPUs.

• The technology offers up to 1.8 TB/s of bandwidth per GPU, which is 14 times faster than the current PCIe Gen5 standard. This means data can move between chips at lightning speed.

• NVLink Fusion can be added to custom chips either through a chiplet or by integrating NVIDIA’s IP directly, providing flexibility for chip designers.

So What?

NVIDIA is no longer just making the chips — it’s laying down the infrastructure for how the future of AI will run.

By opening up NVLink Fusion, they’re setting the rules of engagement and inviting others to build on top of their ecosystem (as long as they play by NVIDIA’s rules).

The message is clear: if you want to build fast, scale smart, and stay in the AI race, better hop on the highway, tolls may apply.

Together with Honeybook

Unlock AI-powered productivity

HoneyBook is how independent businesses attract leads, manage clients, book meetings, sign contracts, and get paid.

Plus, HoneyBook’s AI tools summarize project details, generate email drafts, take meeting notes, predict high-value leads, and more.

Think of HoneyBook as your behind-the-scenes business partner—here to handle the admin work you need to do, so you can focus on the creative work you want to do.

Try HoneyBook for free

Industry Intel

AG-UI Transforms AI Agent Interactions

Last week, we saw CopilotKit release AG-UI, of course not to get confused with MCP or A2A.

Is this a case of another day, another protocol? Maybe. But AG-UI might be the one that actually changes how we interact with AI.

A big question for the future of AI agents was how we’d end up interfacing with them. Until now, most AI agents have been backend operators, quietly automating tasks. Efficient? Yes. Intuitive? Not even close.

AG-UI (Agent-User Interaction Protocol) is an open, event-driven protocol designed to bridge the gap between backend AI agents and frontend applications.

If MCP standardized context. If A2A enabled agent-to-agent collaboration. AG-UI completes the trio - bringing structured, real-time UX into the mix.

Why is this a big deal?

• Developers no longer need custom WebSocket hacks or awkward prompt tricks just to show token streams.

• AG-UI introduces structured event streaming using standard HTTP (SSE). That means live updates, multi-agent handoffs, tool orchestration, and state diffs, all in a clean JSON stream.

• It’s framework-agnostic. You can use LangGraph, CrewAI, or your own setup.

• It ships with SDKs in Python and TypeScript, and supports plug-and-play frontend components via CopilotKit.

So What?

By standardizing the interaction layer, AG-UI makes it easier to create intuitive, real-time experiences without reinventing the wheel each time.

If you’re building user-facing agents, it might be time to look beyond just workflows, and it looks like AG-UI is going to make that accessible for all.

Get Your Hands Dirty!

Create and Animate Stellar Images with THIS Pipeline…

Wait, are you sure you didn’t just crop a video and put it in this newsletter?

Nah. The above is totally AI-generated. With ChatGPT now spitting out hyper-realistic images, I figured, why not put this through Kling to animate it?

Here’s how you can animate any image from ChatGPT:

1. Navigate to ChatGPT and make sure the GPT-4o model is toggled on.

2. Type in your prompt to create an image, example: Create me a scene of a living room looking at a warm fireplace. Use /Create Image to ensure an image will be generated.

3. Save the generated image and go to Kling.

4. Jump to Video Generation on the left sidebar and then hit Image to Video.

5. Scroll down and add a prompt, for example: Animate this so the fire is moving.

🔥 Pro Tip: Use highly detailed image prompts in ChatGPT, movement looks way better when your starting image has strong texture and lighting.

Some sample prompts for you to try:

Create a snowy mountain landscape at dusk with clouds slowly drifting.

Generate a futuristic city at night with neon lights pulsing and flying cars zooming past.

Visualize an ocean scene with waves gently rolling and seagulls flying overhead.

Quick Bites

Stay updated with our favorite highlights, dive in for a full flavor of the coverage!

OpenAI released its new Codex agent in ChatGPT last week to streamline AI-powered software development.

Microsoft launched its Copilot Plus PC and Windows AI efforts last year, and now it’s going a step further today with native Model Context Protocol (MCP) in Windows and the launch of the Windows AI Foundry.

Shanghai-based Synyi AI has launched a trial clinic in the Al-Ahsa region of Saudi Arabia, using AI as the main diagnostic tool.

Databricks last week it would buy database startup Neon in its latest deal valued at about $1B, aiming to strengthen its platform to use AI agents more easily.

There’s bad news for Elon Musk – Grok publicly humiliated him and cast doubt on AI safety.

Trending Tools

📈 virattt / ai-hedge-fund - A proof of concept for an AI-powered hedge fund. The goal of this project is to explore the use of AI to make trading decisions.

🔗 Backlink Management - Automate outreach, monitor links, and boost organic traffic—all from one dashboard using the power of AI.

♾️ n8n MCP Server - Use the MCP Server Trigger node to allow n8n to act as a Model Context Protocol (MCP) server, making n8n tools and workflows available to MCP clients.

The Neural Network

Big news - we just hit 5,000 subscribers!

Massive thanks to every single one of you for reading, sharing, voting, and building with us each week.

You’re the reason Digestible AI keeps growing, and we’re just getting started…

Now about that NotebookLM poll from last edition… We asked:

Would you use NotebookLM to speed up content creation?

It seems like an overwhelming majority of you guys definitely will be doing so…

Until we Type Again…

Thank you for reading yet another edition of Digestible AI. Be sure to give us a follow on X, Instagram, and LinkedIn too!

If you have any suggestions or specific feedback, simply reply to this email. Additionally, if you found this insightful, don't hesitate to engage with us on our socials and forward this over to your friends!