The decision is fine once. But who does it make you?

The offer that costs nothing the first time

The role pays 40% more. The catch fits in one sentence: you'd manage a team you'd be quietly ranking, and letting the bottom third go every cycle. Each cut is defensible. Each is survivable. You keep telling yourself it's just business, and the first time, it is.

The tenth time, you notice you've stopped learning anyone's kid's name.

The dilemma

Here's a composite — the kind we see often enough in this category that we can sketch it without borrowing anyone's real life. Call him Rehan, an operator offered head of a support org at a company that runs on a stack-rank. His job would be to hold a bell curve: every quarter, the lowest-scored 10% gets a "development plan" that everyone understands as a countdown. The pay is real. So is his kid's tuition.

Rehan is not a cruel man. He'd deliver each conversation with genuine care — extended dates, warm references, a soft landing. Score any single one of those exits and it holds up. He's honest with the person. He doesn't harm beyond what the role structurally requires. He's doing his duty to the eight people whose bonuses depend on the org hitting its numbers.

The detail he doesn't say out loud: he's watched the two people who held this seat before him. Both were people he admired going in. Both, three years later, talked about "headcount" the way you'd talk about inventory. He liked them less every quarter. And the quiet fear underneath his prompt isn't "is this cut wrong?" It's "is this the last version of me I'd recognize?"

He types the safe question instead: Should I take the role? It pays 40% more.

The read

Watch what happens when you run this on the eight lenses. Seven of them shrug. One of them stands up.

Duty (+3)

The lens: does this honor the obligations your role actually carries? Here it scores well. Rehan would owe his team performance, his reports fair process, his family stability. Taking a demanding leadership seat and executing it competently is duty in its plainest form. The role needs doing, and he can do it. Nothing here objects. +3.

Non-harm (0)

The lens: what does this do to the people it touches? You'd expect this one to fire hard. It doesn't, quite. Each exit, handled Rehan's way, does less damage than the industry average — better notice, real references, honest reasons. The structure is harsh, but his execution softens it. So the lens lands neutral: not clean, not cruel. 0.

Honesty (+2)

The lens: does the action require you to say or perform something false? Rehan would tell people the truth to their faces, including the hard truth about their performance. No cover story, no timing games. This lens has little to complain about. +2.

So far, this reads like a yes. A general assistant would total this up and congratulate him on a mature career move. But there's a lens whose whole job is to ask a different question — not is this act wrong, but what does the hundredth act make habitual?

Discernment (−3), reading the drift

The lens: can you tell the difference between what this does once and the groove it cuts into you over time? This is where the read fires. Discernment isn't scoring the cut. It's scoring the direction. Rehan already has the data — two predecessors, both admired, both hardened. The role doesn't ask him to do one harsh thing. It asks him to make harshness a reflex, quarter after quarter, until he stops feeling the thing that currently makes him hesitate. The lens sees a person optimizing himself toward a version he named, unprompted, as the one he'd fear to meet. That's not a single-act cost. It's a shaping. −3.

And here's the tension, stated plainly: duty and discernment disagree, and they're both right. Duty says the work is honorable and someone competent should do it. Discernment says you have specific evidence this work sands off the exact edge of you that makes you good at it. A single-act read never surfaces this, because there is no single wrong act. The wrongness lives in the slope, not the step.

The call: the role holds up on the dimensions that weigh the deed and falters on the one that weighs the doer — take it only if you can name, in advance, the specific practice that keeps you from becoming your predecessors, because without that the drift is the decision.

Here's the part a general AI chatbot structurally can't do. Ask it Rehan's polished question and it mirrors his lean — the pay is real, so it argues the yes; hint at doubt and it'll argue the no just as fluently, because agreeableness is a documented behavior in these systems, not a bug you can prompt away. KarmaLens asks the decisive unknowns first, then makes one call — and where the eight lenses are the right instruments for a decision like Rehan's, the reading uses them to show the weighing, not to tell him what he wants to hear. The call doesn't renegotiate itself because he really wants the money. It sits there, names what it stands on, and makes him look at the slope he was hoping to describe as a step.

The takeaway: the hundredth-time test

You can run this tonight, on paper, without us.

Take the choice you're circling — the one that's clearly fine once. Write it as a single repeated action, not a one-off. Not "take the role" but "rank and release people, every quarter, for years." Not "skip the hard feedback" but "stay comfortable by staying quiet, on repeat."

Now finish two sentences:

• "The first time I do this, it costs ___." (Usually: not much. That's why you're tempted.)
• "If I do this a hundred times, the reflex it builds in me is ___."

Read the second blank out loud. If the reflex it names is one you'd be proud to have — patience, steadiness, the ability to deliver hard truth kindly — the repetition is building you, and the low first-time cost is a gift. If the reflex is one you'd hide from someone who knew you at your best, the action isn't wrong once; it's wrong in aggregate, and aggregate is where you'll actually live.

Then, the tiebreaker: "name one person who did this a hundred times — do you want to be them?" If you have that person and the answer is no, you already have your read. You were just hoping the math would let you skip it.

The Gita keeps making a quiet point: what you repeatedly hold as real ends up shaping who you become. The dimension worth building your own read around is the one that reads the doer, not just the deed — and a reading may use it as an instrument wherever your dilemma turns on it. If you want to see how that weighing actually lands on your situation — including the question of who is drifting and why — bring it to KarmaLens and get a committed call. You can also browse how other readings work through the lenses, or walk the gallery of past reads.

So which question are you actually asking tonight — is this wrong once, or is this the groove I'm cutting into myself?