In partnership with
Good morning to everyone who thinks AI labs only buy GPUs.
The Information just walked into OpenAI's racks.
Tens of thousands of Mac minis and Mac Studios.
Today, why those boxes matter.
Talk to your AI tools the way you'd talk to a colleague.
You don't send a colleague a three-word brief. You explain the context, the constraints, what you've already tried. But typing all that into ChatGPT takes forever — so you don't.
Wispr Flow lets you speak your prompts instead. Talk through your thinking naturally and get clean, paste-ready text. No filler words. No cleanup. Just detailed prompts that actually get you useful answers on the first try.
Millions of users worldwide. Works system-wide on Mac, Windows, and iPhone.
Aaron Tilley at The Information reported late last week that OpenAI has bought tens of thousands of Mac mini and Mac Studio machines in recent months.
They come without screens or keyboards. They sit in racks.
And they are there so an AI can practice using a computer the way you do: clicking buttons, opening menus, finishing a task, failing, trying again.
Anthropic is doing something similar.
The report says that lab rents Mac capacity through Amazon Web Services instead of buying the machines outright.
Here is the machinery.
If you want an AI that can run software for you, it has to practice on real screens.
On actual desktops, tousands of them at once.
Same apps. Same menus.
A score for every click that worked.
Something big is coming to Second Order! 🤖
We’re building a new way to help you go beyond understanding the AI revolution and start acting on it.
Your vote will help decide what we build first.
Thanks for helping us build it right. Now, back to it.
A Mac is a good desk for that job.
The CPU, the graphics chip, and the memory share one pool, which helps when the model has to stare at a screen and decide what to press next.
Stack a bunch of headless Macs and you get a floor of desks the agent can train on.
The GPU helps build the model
The Mac is where the model learns to work
High-memory Macs have been hard to get for months.
On the July earnings call, Mac did $10.4 billion for the June quarter, up 29%.
Cook pointed to MacBook Neo and MacBook Pro for that number, so we are not hanging the lab orders on that print.
He also said customers are already using Mac mini for agentic AI and Mac Studio clusters for local models.
So what does this mean for your portfolio?
$NVDA ( ▲ 1.37% ) still owns the big training spend. $AAPL ( ▲ 1.45% ) owns the desk the agent practices on.
As agents leave chat and start operating software, someone has to supply those desks.
Apple already ships the computer everyone knows how to use.
The GPU trade prices who trains the brain. The Mac trade prices who supplies the desk.
What does a floor of Macs change for you?
That’s it for today!
How did you like today’s newsletter?
Second Order is for information and entertainment only and is not financial advice.
We are not licensed financial advisors.
Any tickers or companies mentioned are our opinions, not recommendations to buy or sell.
Do your own research and consult a professional before making investment decisions.