Why local AI matters

Before you install a thing, it helps to know what done actually looks like.

Free path complete = four receipts

By the end of Lesson 4, you'll be able to say all of this honestly:

1. Ollama runs on your machine (ollama --version works).
2. A model is on your disk (ollama list shows your tag).
3. You chatted locally — the Example Ink prompt, in a terminal, with no login screen anywhere.
4. Offline proof — Wi‑Fi off, and it still answers (Lesson 4).

That's not a teaser. That's a working stack you own — and it's worth showing someone.

What comes after (paid — Build Your Jarvis · Module 0)

Membership is Project 1: Build Your Jarvis:

• VS Code + Continue — your builder's workbench.
• Cloud models via OpenRouter (pay-as-you-go, not another £20/mo subscription).
• A prompt library, a system voice, a sense for cost.
• Later: a desktop app, memory in the UI, tools, voice, integrations.

Think of it this way: the terminal chat you build here in Off the Grid is the engine. Build Your Jarvis · Module 0 is where you start designing the car around it — with AI holding the spanners.

Write this commitment (optional, 30 seconds)

My RAM: _____ GB
My model: _____________
Off the Grid goal: run the Example Ink prompt locally before I pay for anything.

Stick it on a note. Lesson 2 starts the install.

And I mean this — free Off the Grid was never a funnel trick. It's the proof I wish someone had made me finish before I went and bought three more subscriptions. Even if you only ever do Ground Zero and Off the Grid and stop there, you've genuinely won.

Continue — the recap.