Naveo

STEP 20 / 24

A7 A/B

MCQ · NO COST

You're 10 turns in trying to get the model to understand your intent. You clarified 3 times. It keeps drifting: uses terms you discarded, assumes things you denied, proposes solutions you already rejected. What do you do?

Why?. optional

Look for: closed contract, explicit fallback, scaffold at the end.

GUEST MODE

You're viewing this lesson as a guest. To save your progress, earn XP, and keep your streak, sign in when you're ready to check.

Sometimes the right button is "new conversation"

The whole first half of the track taught you to sustain a conversation: insist, correct, reaffirm. That works to a point. Past that point, the conversation is contaminated and the best move is to throw it out.

The criterion for detecting the reset moment is practical, not philosophical.

The 3 reset signals

You repeated the same correction 3+ times. If you had to say three times "I don't want JSON, I want YAML", the model is not going to start understanding on the fourth try. previous context is outweighing you.
The model assumes things you explicitly denied. Sign that the weight of "the user probably wants X" beats the weight of "the user said NOT X".
You lost track of state. If you can't sum up in 3 lines what was agreed, the model can't either.

How to start over well

It's not throwing everything out. It's distilling what you learned into a clean prompt:

code

I want <final goal>.
I do NOT want: <discard 1>, <discard 2>, <discard 3>.
I need: <concrete spec>.
Question: <one clear question>.

That recipe loads the useful context of the 10 old turns into 4 lines, without dragging the misunderstandings.

The opposite error: starting over too fast

If you reset every time the model drifts a single turn, you lose the incremental learning. The rule isn't "first doubt, reset". It is:

Situation	Action
First drift	Specific follow-up (lessons 1-9).
Second drift of the same thing	Reaffirm with emphasis (lessons 11-12).
Third drift of the same thing	Summarize and restart conversation.

This is the opposite of the beginner instinct, which is "I already invested 10 turns, it'd be a waste to throw them out". That's sunk-cost fallacy. the 10 turns are already gone. The only thing that decides future cost is how clean you start the next turn.

On the right, the decision moment: keep going or restart?