To do some small-scale model training at work, I eventually recommended buying a Mac Mini. The reasoning wasn’t complicated: I needed it for bioinformatics analysis, running agent deployments, and occasionally training models with modest parameter counts — these scenarios are exactly where M-series chips with unified memory shine. The cost of separate RAM plus a large-VRAM GPU far exceeds the Mac platform; do the math and the choice is clear.
And just like that, I unlocked the achievement of pushing forward multiple devices, multiple platforms, and multiple projects all at once.
Multiple Devices
The devices I’m currently juggling:
- Laptop: Writing documents, making slides, running WSL for light computing and plotting
- Server: Preliminary testing and validation for non-sensitive data
- Mac Mini: Model training, bioinformatics analysis, agent deployment
- Tablet: Remote connection, reading papers, acting as an ad-hoc dashboard
Multiple Platforms
- Linux: Server and WSL — used for bioinformatics and computational tasks
- macOS: Runs on the Mac Mini, though honestly I use it like a server too
- Windows: Still the go-to for daily communication and documentation
- ChromeOS: Every colleague asks, “What system is that? Never seen it…”
Multiple Projects
- Literature Review: Heavy reading and organizing — more of an “input” state
- Algorithm Development: Deep thinking and repeated experimentation — more of an “exploration” state
- Software Development: Engineering mindset and real-world delivery — more of an “output” state
Switching between these three thinking modes is possible, but the switching cost is real.
Too Bad I Don’t Have Multiple Brains
My current state is basically: Async works, but multi-threading doesn’t.
I can only focus on one thing at a time. If I try to parallelize, I end up with every task half-finished, interrupted, and then I have to spend time recalling the context. So-called “multitasking” is just slicing time into thinner pieces.
As I wrote before, agents can efficiently preprocess a lot of the groundwork for me, but tasks still can’t move forward that efficiently — because I can’t check and handle multiple different tasks at the same time.
Maybe This Is the New Normal
For knowledge workers, this “multiple devices, multiple platforms, multiple projects” state is probably going to become more common, not less.
I can’t really say whether that’s a good thing or a bad thing. But since this is the trend, instead of complaining, I’d better find ways to adapt.
At least, let me start by finishing this blog post.
