
My wife and I bought the 11-inch iPad Air M3, and the first thing we both noticed was the weight. It feels noticeably lighter than carrying a full laptop, and that one difference has already changed how often she reaches for it. We bought last year's model because it was the right value for what she actually needs, not because we were trying to optimize for top-end specs.
Why We Chose This Setup
Her main use case is studying, including prep for an upcoming machine learning course at Stanford. For that kind of work, she needs something fast to open, easy to carry, and comfortable for long reading and note-taking sessions. She does not need local active development on the device, so paying for a full laptop workflow would have been unnecessary overhead.
When she wants to experiment with OpenClaw, she can still do that through a remote machine running on a DigitalOcean droplet. That split works well: the iPad handles interface, reading, writing, and navigation, while heavier compute or tooling stays remote.
How the Setup Came Together
The accessories ended up mattering more than I expected. A Bluetooth keyboard and an inexpensive $20 case with a built-in stand made it stable for longer desk sessions. The setup is still simple, but it now covers the basics without friction.
The Apple Pencil Pro also worked better than expected for her. Compared with an e-reader stylus, writing feels smoother and more responsive, and the higher refresh rate reduces the slight lag she notices during note-taking. On paper that difference sounds minor, but in daily use it makes handwriting feel more natural.
Where It Fits Her Workflow
For her current workflow, the iPad Air covers the core tasks: reading, annotating, writing, joining calls, and light remote workflows. The key trade-off is that it is not trying to be a local development machine, and that boundary is what makes it practical day to day. By being explicit about scope, we avoided overbuying and ended up with a tool she uses every day.
Right now, she sees it as a decent laptop replacement. I think that is the useful framing: not whether it can replace every laptop, but whether it supports the work you do most days.
Related posts
Short Review: Deep Learning Part I
February 7, 2026
Thoughts on the mathematical foundations section of Goodfellow, Bengio, and Courville's Deep Learning textbook.
Making Responsive Images Just Work
March 4, 2026
Instead of manually managing -sm/-md/-lg assets, I moved to a manifest-driven workflow that reduced duplication and made performance outcomes more consistent.
Everyone Is a Builder Now
February 28, 2026
Over cocktails, my colleagues and I talked about why AI lowering the barrier to building is a net positive.
