I stopped treating AI cost and doomscrolling as separate problems (solo Mac dev notes)

Source: Dev.to

Background

For months I thought I had two unrelated issues:

1. My AI coding costs were all over the place.
2. I was doom‑scrolling.

It turned out both stemmed from the same problem: attention fragmentation.

The Insight

When I got distracted, my prompts became sloppy, leading to higher token usage and more retries. By tracking spend per session alongside focus quality, I noticed a clear pattern:

• Expensive sessions coincided with broken attention.
• Doom‑scrolling days resulted in higher AI spend.
• Deep, uninterrupted blocks lowered cost per shipped task.

That framing shift changed everything.

Tools

• TokenBar – a live token/cost counter in the menu bar.
  ($5)
  Seeing the cost in real time prevents the “I’ll clean this up later” lie.

• Monk Mode – removes feed surfaces while keeping utility parts of apps.
  ($15)
  This stops infinite feeds from hijacking a build window.

New Workflow

Before launching an AI run, I write a short plan:

• Goal
• Done condition
• Max retries
• Fallback model

This simple structure cuts panic loops and yields:

• Fewer late‑night “what did I even ship?” sessions
• Fewer expensive retries
• Better output quality from shorter, cleaner prompts
• Less emotional thrash

Takeaway

There’s no magic—just better feedback loops. If you’re a solo builder, don’t separate productivity from cost control; they’re the same system.