Skip to main content

Local Clarity > Global Commitment

Introduction

From the outside, my behavior can look inconsistent.

One month, I’m deeply focused on Project X. I talk about it as if it’s a multi-year commitment.

Then something shifts.

Now I’m working on Project Y.

To observers, this can look like:

  • lack of discipline
  • lack of follow-through
  • changing priorities arbitrarily

But internally, it’s not arbitrary at all.

What’s actually happening is:

I continuously re-rank my priorities based on a multi-factor evaluation function.


The Evaluation Function I’m Using (Implicitly)

At any given moment, I’m evaluating each project along several dimensions:

1. Time to Meaningful Progress

  • How long until I can produce something non-trivial?
  • Can I make visible progress in a single session?
  • Is this a short ramp or a long ramp?

2. Types of Upside (Not Just One)

I don’t think about value as a single dimension.

I’m implicitly weighing multiple types of upside:

  • Compounding upside (long-term leverage)
  • Immediate output (content, code, artifacts)
  • Learning / capability gain
  • Positioning / optionality
  • Network / exposure
  • Enjoyment / intrinsic pull

A project doesn’t need to dominate every category — but it needs to score high enough across a few of them.


3. Clarity

  • Do I know exactly what the next step is?
  • Is the path obvious, or still branching?
  • Do I understand what “good” looks like?

4. Friction / Re-entry Cost

This is one of the most important variables.

  • How hard is it to resume this work?
  • Do I need to reload a large amount of context?
  • Is the state captured or lost?

This is also where my systems matter.

→ See: The Zero Overhead Cold Start Habit

Low-friction re-entry allows me to:

  • switch contexts without penalty
  • pause without losing progress
  • maintain multiple active threads

Why My Priorities Can Shift Suddenly

The key point is that this evaluation is not linear.

It can change abruptly.

For example:

  • A project that felt clear becomes ambiguous
  • Another project becomes trivial to execute
  • A new idea scores high across multiple upside dimensions
  • Re-entry cost increases due to lost context

When that happens:

The ranking of my projects can reorder instantly.

There’s no gradual transition. Just a different answer to the question:

“What is the highest-value thing I can do right now?”


Local Optimization vs Long-Term Commitment

Most people optimize for consistency:

  • Pick something
  • Stick with it
  • Avoid switching

I’m optimizing for something different:

At each moment, I want to allocate effort to the highest-scoring option.

That means my system is:

ApproachBehavior
Commitment-basedStable over time
Evaluation-basedContinuously adaptive

Why This Only Works If Switching Cost Is Low

This approach breaks down if switching is expensive.

If every context switch requires:

  • rebuilding state
  • rereading everything
  • re-deriving intent

Then I would just thrash.

So this only works because I’ve invested in:

  • capturing state
  • structuring work
  • minimizing re-entry cost

Example: fmc and Time-to-Value Tradeoffs

A concrete example of this is how I approached Front Matter Canonicalizer.

In 2024, I intentionally didn’t fully automate it.

At the time, the evaluation looked like:

  • high implementation cost
  • low immediate return
  • useful manual workflow already exists

So I kept it minimal and used it as an audit tool.

As I wrote here:

“More importantly — it’s not worth the time investment right now. The cost of building 100% conformity and logic-based automation outweighs the value I’d get at this stage. I’m more focused on getting things to a solid baseline.”

Later, in 2026:

  • the cost of building tooling dropped significantly (thanks Claude Code!), so implementation became trivial, in a few days I got at least a month's worth of work done.
  • the upside increased, so the value of having a way to do mass CRUD (create, read, update, delete) increased.
    • I had way more documents, traffic to site was starting to meaningfully increase
    • All this metadata can be fed into downstream tools that I have built in the meantime (thus further increasing the value of this automation)

So I revisited it and fully expanded it.

From the outside, that might look like inconsistency.

Internally, it was:

The evaluation function changed, so the decision changed.


Creative vs Mechanical Work

Another pattern I’ve noticed:

  • Mechanical work → I fully automate
  • Creative work → I keep manual control

That distinction feeds directly into how I prioritize:

  • If something is repetitive and low-value → automate
  • If something benefits from context and judgment → keep it manual

Why This Looks Irrational to Others

Observers tend to assume:

  • priorities should remain stable
  • plans imply commitment
  • switching requires justification

What they don’t see:

  • changes in clarity
  • changes in time-to-progress
  • changes in upside
  • changes in friction

So the behavior looks like:

“You said X mattered. Now you’re doing Y.”

But what’s missing is:

The internal re-evaluation that happened in between.


Common Misinterpretation

This pattern often gets labeled as:

  • “shiny object syndrome”
  • lack of discipline
  • inability to commit

But the real distinction is:

BehaviorReality
Random switchingNo evaluation
Structured switchingContinuous re-ranking

Practical Heuristic

When I shift priorities, the useful question is:

What changed in the evaluation function?

  • Did clarity improve somewhere else?
  • Did ambiguity increase here?
  • Did the upside shift?
  • Did friction increase?

Closing

From the outside, this can look inconsistent.

From the inside, it follows a simple rule:

At any given moment, I allocate effort to the highest-value, lowest-friction, highest-clarity opportunity available.

That leads to:

  • sudden shifts
  • non-linear progress
  • changing priorities

But underneath it:

There is consistency — just not the kind that is easy to observe externally.