Sunk Cost — Operating Your Human

The machinery treats what it has already spent as a reason to spend more. This is exactly backwards, and it is one of the most expensive errors the system runs.

A sunk cost is a resource already gone — time, money, effort, years — that cannot be recovered no matter what happens next. Sound operation would ignore it entirely; the only question that should govern the next decision is whether continuing is worth it from here. But the apparatus can’t ignore it. The size of the prior investment exerts a pull all its own, and the larger the amount already poured in, the harder the system fights to keep pouring — not because continuing is wise, but because stopping would mean the prior spend was for nothing.

The Quitting entry covers the leaving. This entry covers the trap that keeps the operator from leaving when they should.

WHY THE SYSTEM CAN’T LET GO

Two mechanisms drive it.

The first is loss-aversion. The system reads the abandonment of an investment as a loss, and loss registers harder than equivalent gain. Walking away from the half-finished project, the long relationship, the years in the wrong career, the system processes as losing everything already put in — even though that everything is already gone regardless. So it clings, to avoid feeling a loss it has, in fact, already taken.

The second is the threat to the self. To stop is to concede that the prior investment was a mistake, and the apparatus defends past decisions the way it defends the body, because they’ve fused with identity. Continuing protects the story that the operator chose well. The Ego entry covers that defense. Here it produces a specific distortion: the operator keeps funding a failing course to avoid admitting they funded it in the first place.

Both mechanisms point the same way — throw good resource after spent resource — and both feel, from inside, like commitment, loyalty, or grit. That is the disguise. The Persistence entry covers the version of staying that’s genuinely worth it. Sunk-cost staying wears persistence’s clothes while doing the opposite work.

THE HOW — RUNNING THE DECISION CLEAN

The correction is a single reframe, applied deliberately, because the system won’t apply it on its own.

Ask the question that deletes the past from the math: knowing only what I know now, with the prior investment already gone and unrecoverable — would I start this today? Not “have I put a lot in.” Not “would it be a waste to stop.” Would I begin, from here, given the current odds and current costs. If the honest answer is no, the only thing keeping the operator in is the spend they can’t get back anyway.

To run it concretely: separate the two ledgers. On one side, everything already spent — write it off; it is gone in every scenario, including continuing. On the other side, only what remains to be spent and what it might still yield, from this point forward. Decide using the second ledger alone. The first one feels relevant. It is not. It is history, and history does not get a vote on the next move.

To catch the trap forming, listen for the tell in the system’s own reasoning. When the justification for continuing is mostly about the past — “I’ve already put in so much,” “after all this time” — rather than the future returns, the sunk cost is steering. Sound reasons to continue point forward. Sunk-cost reasons point backward.

THE LANDING

None of this means every long effort should be abandoned at the first doubt. Plenty of investments are worth completing on their genuine forward merits, and the Commitment entry covers staying for the right reasons. The point is narrower: the amount already spent is never one of those reasons.

What’s gone is gone. It will be gone whether the operator continues or stops.

The only resource the one in the chair still commands is the part not yet spent. Spend that on where things stand now — not on rescuing a past that no decision can reach.