A deterministic thinking instrument
Don't just read the answer.Read what it's doing.
Paste an AI answer, report, proposal or argument. Unthinkit reveals explicit patterns in the language, links each observation to the evidence, and shows you a thinking move to try next.
How it works
From answer to evidence to judgement.
- 01
Detect
Nine deterministic lenses look for explicit patterns in the answer.
- 02
Link
Every finding points back to the exact words that triggered it.
- 03
Look again
Each pattern becomes a thinking move to try next, not a verdict on the answer.
- 04
Judge
You decide whether the observation matters and what to do with it.
The nine lenses
Nine ways to look again.
Each lens looks for explicit language patterns. They are not ranked, scored or treated as errors. They are prompts to look again.
- 01
Reframing
Spots when the answer changes the frame — then helps you compare what each framing reveals or hides.
- 02
Compression
Spots when a complex issue is simplified — then helps you restore missing detail, dependencies or exceptions.
- 03
Contradiction
Spots when claims are presented as incompatible — then helps you separate and test them.
- 04
Evidence
Spots when support, proof or verification is in question — then helps you ask what evidence would actually matter.
- 05
False Precision
Spots exact-looking numbers wrapped in approximate language — then helps you test whether the precision really matters.
- 06
Tension
Spots trade-offs between competing considerations — then helps you name what is being balanced.
- 07
Scale Shift
Spots shifts in level, timeframe or geography — then helps you test whether the conclusion transfers.
- 08
Surprise
Spots language of unexpectedness — then helps you identify the hidden expectation or baseline.
- 09
Vagueness
Spots imprecise scope, quantity, duration or magnitude — then helps you make the statement more specific.