← All newsletters

The iPad Mini 2030 Life, and How It Works

Nov 18, 20266 min readNewsletter #12

The iPad Mini 2030 Life, and How It Works
Text size

Remember the iPad Mini 2030 life? Here's how that works

Fun fact: an iPad mini is 300g, which is 4x lighter than your 1.2ish kg MacBook Air.

I love having a light iPhone with nothing except good old fashioned iMessage and call. No social media, no work apps. No ebooks. No podcasts. No excuse to be in the mind when I don't need to be.

Which makes my iPad Mini the work horse:

  • A bigger screen than an iPhone Pro Max
  • Easier to carry because I put it in my backpack
  • And no buzzing or weight in my hand because it’s light and silent when in my bag

Not sure about you, but I don't think I will ever buy the iPhone fold.

In 2030, the work life balance is in what devices you choose and what you choose to use them for.

And you will feel like Tony Stark minimalist. When people see you running Claude Code on an iPad mini, with a tiny stand and a keyboard at a coffee shop... people will ask if you're a monk who writes. It’s like wearing Uniqlo, light and breezy without all the pretentious fuss (not sponsored).

Now your Second Brain OS via Claude Code doesn't run on the iPad. It runs on your Mac. Your knowledge is on the Mac. Claude Code is on the Mac. The iPad just needs a way in.

That way in is called SSH. Secure Shell. It's how developers have been reaching computers on the other side of the world since before Google existed. You type on the small device, the Mac does the work, the screen comes back to you.

Just a clean connection between your iPad Mini and your Mac, commanding it in real time.

Now, a confession, because most people selling this are quietly lying about it.

Claude Code talks to Anthropic's model no matter how you run it. Terminal, VS Code, the desktop app, the browser — the model is in the cloud in all four. If someone tells you their setup means your AI never touches the internet, they're describing a local model, which is a different product at a different quality.

So why the terminal? Because the terminal is the only version that isn't missing pieces.

So why the terminal? Because by now, your Mac isn't running a chat window. It's running an environment.

Every program on it is a tool your AI can use. Merge the contracts. Compress the photo. Pull the text out of forty PDFs and search all of them for one clause. Query the spreadsheet Excel refuses to open. Transcribe a client call without the recording ever leaving the machine.

That's what an hour of setup bought you. And it's worth more than the laptop it's sitting on.

Which brings me to the thing people get wrong. They assume an environment like that chains them to the desk it lives on.

It doesn't. The work happens on the Mac. The Mac can stay home. What you carry is a window onto it, and the window can weigh 300 grams.

Not a cut-down version. Not an app with three of the features. Every command, every skill, every indexed note — running on the machine in your study while you're holding something the size of a paperback.

And sit with that for a second, because it's better than it first sounds. The heavy work and the heavy device stopped being the same object. A machine at home can be plugged in, cooled, and left running for an hour while you go and do something else. Carrying it was never the advantage. It was the tax.

(The app versions can't stand in for this, by the way. Anthropic's own docs say the VS Code extension runs a subset of the commands — their recommended fix is to open a terminal and run claude there. And the browser version doesn't run on your machine at all; it clones a GitHub repo into Anthropic's VM, so untracked files never arrive. Your reflections folder isn't a GitHub repo. It's a folder of notes on a disk in your house.)

P.S. Here's why you SHOULDN'T follow a YouTube tutorial.

Problem one: the front door.

The tutorial says 'open port 22'. A port is a numbered door on your Mac for network traffic, and there are about 65,000 of them. Port 22 is the one the world knows to knock on.

So you open it, and congratulations, your Mac is now a house on the internet with a sign that says 'try passwords here'.

I'm not being dramatic. Leave that door open for one evening and check the logs in the morning. Hundreds of attempts. From Romania, from a data centre in Ohio, from a fridge in Shenzhen that someone forgot to update. They're not after you. They're after every open door on earth, and a script does the knocking while they sleep.

Your knowledge folder — the emotions, the observations, the things you'd never say out loud — sitting behind a password a bot might eventually guess.

The fix isn't a better password. It's not having a door on the public internet at all. Make it impossible for the hackers, creeps, strangers and demogorgons to find you.

Problem two: the connection that dies and BREAKS your work with it.

Imagine that you've connected from your iPad Mini to your Mac at a coffee shop, Claude Code is 20 minutes into setting up your new beautiful website and the wifi disconnects. RANDOMLY!

Connection breaks. And the terminal session dies with it over standard SSH.

Not pauses. Dies. Like the lights going out mid conversation and your memory vanishes when they come back. 'What were we talking about?'

Because out of the box, the remote session belongs to the connection. No connection, no session, no Claude Code. You reconnect, and get a blank cursor asking you what you'd like to do today. As if nothing ever happened.

Same thing when the iPad locks its screen and goes to sleep. Same thing when you switch from WiFi to 4G in the car park. Same thing when the battery dies.

And the fix is the same tiny trick developers have used on servers for 30 years: session persistence.

The result? you're not 'running' Claude Code from the iPad. You're peeking through a window at Claude Code running on the Mac. Close the window, the session is still there. Open it again and pick up from where you left, hassle free.

Set both up wrong and you'll do what everyone does: go back to carrying your Macbook Air or always working from home because you have a Mac Mini (BURN!). Set both up right, once, and you never think about either again.

P.S. Warning: it's addictive to use an iPad mini to SSH into my Mac at 3 am when I get my best ideas. Which is why I leave it in the bag pack (mostly!).

The Productivity Coach does it right. $198, once, one afternoon. It's the second thing I'd buy after the Knowledge Work Coach, and I'd buy it before anything else in this letter.

It also reads your week back to you, every day, week and month, and puts the week's tasks in your Todoist. That's a letter of its own. Here are the three things it does for the iPad life:

A private tunnel. Between your devices and nobody else's. Your Mac is invisible to the internet and visible to your iPad and iPhone. The door opens from inside your house and no rando/creep/stranger/hacker can enter.

A Claude Code session that never dies. It lives on the Mac whether you're connected or not. Lid closes, internet disconnects, battery dies or your iPad gets stolen. You reconnect and Claude Code is exactly where you left it. Mid sentence with the words you typed but didn't enter still there.

Your backup, if you want one. It offers to put your folder in a private GitHub repository (your call, these are your reflections), and then makes sure you can back up with 'git push' after you 'git commit' directly from your iPad. That's the part that quietly breaks when you set it up yourself: it works on the Mac and fails from the iPad. So that you don't have to go back to your Mac to back up to the cloud.

[Screenshot: an iPad Mini, the terminal open, Claude Code mid task on the Mac at home]

That's living 2030. A MacBook Air at home doing the work, and you at the beach, the café, the school run, or in bed, with a screen the size of a paperback.

Kevin, that's the iPad Mini. Not the Mac Mini. Read it again.

That was Part 4 of 10.

Next: Should Any of Your Second Brain Face the Public?

Keep reading