The Hyperfocus Paradox: Building Side Projects with a Neurodivergent Mind • James Best

I’m back after a few weeks/months away. Between interviews, school holidays, and the usual chaos, writing took a back seat. But that’s not the full story.

In the last week I’ve been completely consumed by side projects. Breedr Devtools and IDA (Intelligent Digital Assistant) have taken over every spare waking thought. If you’re neurodivergent, you’ll recognise this pattern immediately—once an idea takes hold, putting it down becomes nearly impossible.

This is something I’ve struggled with for years: the inability to schedule learning when hyperfocus strikes. When IDA appeared in my mind, everything else became a write-off outside of day-to-day work. It’s a pattern that’s both incredibly productive and frustratingly unpredictable.

The Nature of Hyperfocus

Hyperfocus isn’t simply “being really interested in something.” It’s a state where your entire cognitive capacity narrows to a single point. Time becomes fluid. Meals get skipped. Everything else fades into background noise.

For me, it manifests most strongly with side projects. A small idea—perhaps a developer tool that could save a few minutes here and there—suddenly demands my complete attention. What starts as a curiosity becomes an obsession.

The challenge isn’t the focus itself. The challenge is that it arrives unannounced and overstays its welcome, disrupting any carefully planned learning schedule I’ve set up.

Working With the Current, Not Against It

I’ve tried fighting hyperfocus. Forcing myself to stick to predetermined learning goals whilst my brain screams about this one particular problem that needs solving. It doesn’t work. The result is neither good learning nor good building—just frustration.

What’s working better is accepting these cycles as part of how I operate:

Recognise the signs early

When I find myself thinking about a project during completely unrelated activities, that’s the signal. My brain has latched onto something. Fighting it will be exhausting.

Set minimal boundaries

I can’t stop hyperfocus, but I can contain it slightly. Sleep remains non-negotiable. So does time with family. Everything else becomes flexible.

Trust the ebb and flow

Hyperfocus eventually releases its grip. Sometimes after a few days, sometimes weeks. The project either reaches a natural pause point or my interest shifts. Crucially, this isn’t failure—it’s completion of a cycle.

Capture the output

During hyperfocus, I’m extraordinarily productive. The trick is ensuring that productivity translates into something useful. For IDA, that meant 33 commits in a week: OCR functionality, improved screen capture, database migration to libsql. These aren’t half-finished experiments—they’re working features.

Keep a backlog for the quiet periods

When hyperfocus fades, there’s often a lull. This is when structured learning works well. I maintain a list of courses and topics that I can pick up during these calmer phases. Matt Pocock’s Vercel AI SDK course is sitting there ready for when my mind quiets.

The Role of Support

My wife has learnt to recognise when I’m in one of these sessions. She doesn’t try to pull me out—she knows it wouldn’t work. Instead, she helps maintain the boundaries that prevent complete burnout. It’s not something I could manage alone.

If you’re in a similar situation, finding people who understand these patterns makes an enormous difference. They don’t need to be partners—colleagues, friends, or communities of other ND folks can provide the same scaffolding.

Making Peace With the Pattern

I used to view hyperfocus as something to overcome. A bug in my operating system that needed fixing. Now I see it differently.

Yes, it makes traditional planning difficult. But it’s also responsible for some of my most valuable work. Breedr Devtools emerged from one of these cycles—65 commits delivering features, bug fixes, and significant refactoring. Without hyperfocus, it might have remained a vague idea that never materialised.

The goal isn’t to eliminate these intense periods. It’s to create enough structure around them that they become productive rather than destructive.

Practical Adaptations

Here’s what’s working for me:

Flexible learning schedules: Rather than “Learn X every Tuesday,” I maintain a queue of topics I’m interested in. When focus allows, I pick from the queue.
Project rotation awareness: I know most projects will have their moment and then fade. That’s fine. The goal is progress during the active phase, not perpetual maintenance.
Version control as memory: Everything goes into Git immediately. When I return to a project weeks later, commits tell me what past-me was thinking.
Minimal viable documentation: During hyperfocus, writing docs feels impossible. So I write just enough: a README with setup steps, inline comments for complex logic. Future-me is grateful.
Energy monitoring: Some days hyperfocus is productive. Other days it’s spinning wheels. Learning to recognise the difference prevents wasted effort.

The Broader Lesson

This isn’t just about neurodivergence. Many engineers experience something similar—that pull towards building something that won’t let go until it’s done.

The pattern I’ve described works for me, but the underlying principle is universal: work with your natural rhythms rather than against them.

Traditional productivity advice often assumes everyone operates the same way. Consistent daily habits. Steady progress. Predictable output. For some people, that works beautifully.

For others, productivity looks different. Intense bursts followed by quiet periods. Deep focus on one thing rather than shallow engagement with many. Neither approach is superior—they’re just different operating modes.

Looking Ahead

I’m not declaring victory over the challenges of hyperfocus. Next week, another idea might arrive and derail everything I’ve planned. But I’m getting better at recognising the pattern and working within it rather than fighting it.

If you’re struggling with similar patterns, the most important thing I’ve learnt is this: you’re not broken. Your brain works differently, and that difference can be a strength if you structure your environment to support it.

The key is finding what provides just enough structure to be productive without so much rigidity that it becomes impossible to maintain. For me, that’s flexible learning queues, project rotation awareness, and support from people who understand the pattern.

Your optimal structure will likely be different. The important part is giving yourself permission to discover it rather than forcing yourself into a mould that doesn’t fit.

Now, if you’ll excuse me, I have some ideas about IDA’s next feature that won’t let me think about anything else.