AI Unplugged · by Mike Murphy · sent August 28, 2026 · Issue 073
Nvidia's $13 Billion Hug 🤗
Nvidia is reportedly moving to acquire Hugging Face, the open-model hub nearly every AI developer relies on, for $12.9 billion — a bid to extend its grip over the open AI ecosystem.
🌲 Agent Evergreen Build Log — Week One
The core idea behind this project is that the messy notes you take while building something are the real teaching material, if you can capture them without stopping to perform.
This week I proved that out for the first time. I took my own rough build notes and had an AI agent turn them into the first real section of the system’s instruction manual, almost unchanged. Small step, but it’s the first evidence the pipeline actually works.
- Waitlist signups: 2
- Read the full manifesto: mikemurphy.ai/articles/agent-evergreen-manifesto
- The founding waitlist: mikemurphy.ai/evergreen
🗞️ AI News of The Week: Nvidia to Acquire Hugging Face for $13 Billion 🤗
Nvidia is reportedly moving to buy Hugging Face, the open model repository nearly every AI developer relies on, for $12.9 billion.
Hugging Face works like GitHub for AI models: developers and researchers search for models that fit their needs, download and run them, fine-tune them, and upload the new versions back for others to use.
Why this matters:
- Infrastructure grab: Hugging Face has become the default place to store and download open-weight models, giving Nvidia real influence over that ecosystem
- Hedge against frontier labs: OpenAI, Anthropic, and others are increasingly building their own custom chips, so owning Hugging Face gives Nvidia a way to keep the open-model side of the industry leaning on its hardware
- A second shot at AI cloud: Nvidia has struggled to get its own cloud AI business off the ground, and this could help restart that effort
- Reach into robotics: Hugging Face has been investing in models for robotics and other physical AI applications, an area where Nvidia is already a major player
👉 Read More About the Nvidia-Hugging Face Deal
🧠 AI Term of The Week: Dogfooding 🐶
Definition: Dogfooding is when a company uses its own product internally, in real work, before or while selling it to customers.
Explain Like I’m 5: It’s like a chef tasting their own cooking before it goes out to the table.
Why it matters now:
- AI tools change so fast that using them yourself, daily, is often the only way to catch what’s actually broken
- Companies that dogfood tend to ship fixes faster because they feel the pain first
- If a company won’t run its own business on its own AI tool, be wary
💬 Quote of the Week
“Wisdom is not a product of schooling but of the lifelong attempt to acquire it.” — Albert Einstein
🧰 AI Tool of The Week: Handy 👋
Handy is a free, open source speech-to-text app that lets you dictate straight into any text field on your computer.
Why I use it:
- Honestly, probably because of the fun, squishy logo, but it’s also a good, free, open-source alternative to WisprFlow, which is awesome, but not free.
What it does:
- Turns your voice into text in any app, not just one specific chat window
- Runs quietly in the background, triggered with a hotkey whenever you want to talk instead of type
- Free and open source, no subscription required
🛠 AI Tutorial Of The Week: Astro: How to Build Your First Website
Description: This tutorial walks you through building your first website with Astro, a framework built for fast, content-driven static sites, using AI to help rebuild a real site from scratch.
🤖 What You’ll Learn
- Scaffolding a new Astro project with one command, and picking between a minimal and basic starter template
- What goes in the public folder versus the src folder, and how to add your first .astro page
- Running the local dev server with npm run dev, including how to start, stop, and restart it
- Using AI to rebuild a real site from your own assets and a simple prompt
Mike Murphy — The AI Handyman 🧰 Helping creators & small businesses turn their content & documents into AI-powered tools.