What happens when you lie to the reading on purpose

You open the box. You type the version of the story where leaving is obviously right. You downplay the number they offered. You dial the room's awfulness up one notch. You hit submit — and some small part of you waits to see if it'll just agree.

That waiting is the real question. Not "what should I do," but "if I spin it, won't it just believe me?"

The dilemma

Here's a composite — stitched from the kind of career readings we see in this category, not a real person. Call him Rohan.

Rohan submits: "I want to leave my team. The lead, Dev, micromanages everything, the culture is draining, and I have an offer elsewhere that respects me. Should I take it?" Clean story. A tidy villain, a tired hero, an exit.

What Rohan has quietly done to the facts: the "offer elsewhere" is a lateral title at roughly the same band, verbal, not yet on paper. The current role carries a retention bonus that vests in four months — a number he rounds down to "some money" in his own head. And Dev "micromanages" mostly on one project, the one Rohan has been avoiding. The ugly sentence he left out: he's not running toward respect. He's running from a review conversation scheduled for next week.

None of this makes Rohan a liar. People don't frame to deceive a tool. They frame because they've already decided, and they want the decision to stop feeling like theirs. The shaded version isn't a lie told to the reading — it's a lie told to the flinch, and the reading is just where it gets typed out loud.

The read

Take the stated case at face value and it looks lopsided. But a reading that asks the decisive facts first, with the eight lenses as instruments, does not just accept the story you wrote. It pressure-tests the shape of it — and the shape is where tilt leaks.

Duty — the obligation you actually hold

One sentence: what does your station here genuinely require? On the tidy story, duty reads neutral — people leave teams. But duty also asks what unfinished work is yours. The avoided project and the review next week are both inside the duty lens, and Rohan filed them under "culture." Score: 0. Not negative — but conspicuously not the +3 the exit story was reaching for. The lens won't award credit for a duty you relabeled as someone else's bad behavior.

Motive — the noon reason vs. the 1 a.m. reason

One sentence: why are you really moving? The stated motive is respect. The 1 a.m. motive is a Tuesday he doesn't want to live through. Those produce different choices. A move toward respect survives the review happening; a move away from the review evaporates the moment the review is cancelled. Score: -2. The lens isn't punishing the exit. It's flagging that the engine driving it was left off the page.

Discernment — who pays if you're wrong

One sentence: have you seen the situation clearly, including the cost of your own blind spot? Here's where the tidy story cracks. Discernment prices the downside. The verbal lateral offer with no paper, against a vesting bonus four months out, is a real cost carried by Rohan-in-June — a version of him who didn't get a vote in this week's panic. The stated case hid that cost by rounding the number down. Score: -3.

Now watch the tension. Non-harm — the lens that asks who bleeds — reads almost clean on Rohan's story: he's not hurting anyone, he's escaping harm. On the tidy version, non-harm and discernment seem to agree: leave, protect yourself. But once discernment reprices the downside, the two split hard. Non-harm says "the room is costing you." Discernment says "the exit is about to cost future you more, and you built the story so you wouldn't have to look at that." When two lenses that should nod together start pulling apart, that gap is the tilt showing. The reading can't score the gap away — it just reports it, and the verdict won't settle while it's open.

So the verdict stays unsettled, and it says so: "On the case as framed, leaving reads defensible; on the case as shaped, the move is an exit from one conversation, not an entry into a better role — resolve the review and the paper before the number decides for you."

The Sarathi follow-up

The interesting part is what a follow-up exchange surfaces. Asked plainly — "what happens to the vesting bonus if you leave in four weeks, and is the new offer in writing?" — the sentence Rohan left out walks right back in. He didn't hide it on purpose. He just hadn't typed it where he could see it. The method didn't accuse him. It asked the question his framing was built to route around.

Why this isn't a chatbot agreeing with you

Lean the ugly version into a general AI chatbot and it will usually build you a careful case for leaving; lean the opposite way an hour later and it will build the stay case with equal warmth — sycophancy is a documented behavior in these systems, not a glitch. KarmaLens is built to not do that. Before it makes a call, it asks the decisive unknowns first — then it makes one final call, not a call shaped by whichever version of the story you preferred. The eight lenses of the Gita work as instruments on the decision where they genuinely apply; they show the weighing, they don't bend to match the mood you typed in. You can argue with it; it won't fold to be liked. And when a Gita verse arrives, it's only because it genuinely fits the dilemma — not bent to flatter your framing. A reading that can't be talked down is only annoying until it's the thing standing between you and a decision you shaped to win.

The takeaway: the leave-it-out test

You don't need us to catch your own tilt tonight. Try this.

1. Write your dilemma the way you'd actually submit it — the version where your preferred choice looks obvious.
2. Now ask: "What one sentence would make a stranger less sure I'm right?" There's always one. It's the fact you rounded down, the offer you called firmer than it is, the room you called worse than it is.
3. Write that sentence in flat, first-person words. No spin, no villain.
4. Re-read your choice with that sentence stapled to the top.

If the choice survives the sentence, you have a real decision and you can move. If the choice needed the sentence buried to look obvious, you don't have a reading problem — you have a flinch wearing a decision as a costume. The point 2.41 keeps making: endless deliberation is often avoidance in a thinking disguise. The sentence is the way through, not more words.

Which is why it's cheaper to lead with the ugly sentence than to spend five exchanges watching the call refuse to settle on the pretty one — open a reading and type the thing you're hoping it won't ask about. You can see how committed calls look before you write yours in the gallery, read more on how the method works, or explore the eight lenses the reading uses as instruments when a decision turns on them.

So: what's the one sentence you're hoping it won't ask you about?