Directional improvement

A commitment to move a valued variable in a better direction without pretending you know the optimal endpoint.

Somewhere in your forties the question "what do I actually want my life to be about?" started recurring, and every attempt to answer it with a grand destination has collapsed within a month. That's not a personal failing. The endpoint is genuinely unknowable from here, and pretending otherwise produces goals you abandon or fake.

Directional improvement is the destination form for open terrain: a commitment to move a valued variable in a better direction from a baseline, without pretending you know the optimal endpoint. Use it when the domain is open-ended, the endpoint is contested or unknowable, and feedback is available. The spec requires five elements to keep it honest: a direction and baseline (more of this month in activities producing both vitality and credible service, compared to last month); evidence that movement is genuine, not just mood; floors and anti-goals; a review cadence; and an "enough for now" or stop condition for the current cycle. What changes: you get real steering without a fake finish line, and each month you can check whether the needle moved and inspect what caused the shift. The trap the spec names directly: "improve" as a permanently unfalsifiable aspiration, with no magnitude, no horizon, and no point at which attention should move elsewhere.