Skip to content
Substance over spectacle

The science, plainly

Walni is built on a simple discipline: measure carefully, train adaptively, and explain every recommendation.

ReactionAttentionInhibitionMemorySpeedTimingCross-modalEpisodicSpatialExecutive

A sample cognitive profile

Ten domains, one score each — this is the same radar view you'll see in your own dashboard after your first sessions. Values shown here are illustrative, not real data.

What we measure

Reaction time

How quickly you detect and respond to a simple stimulus.

Sustained attention

Holding focus and responding accurately to relevant cues over time.

Response inhibition

Withholding an automatic response when the situation calls for restraint.

Working memory

Holding and updating information in mind across a short span.

Processing speed

How fast you can accurately scan and match visual information.

Temporal processing

Perceiving and reproducing timing, rhythm, and duration accurately.

Cross-modal integration

Binding what you see and what you hear into a single, coherent response.

Episodic memory

Encoding and recalling specific experiences — what happened, and where.

Spatial reasoning

Understanding depth, rotation, and the arrangement of objects in space.

Executive function

Planning ahead, sequencing actions, and adapting strategy under pressure.

AI is quietly changing how we think.

Three peer-reviewed studies, 2025 — pointing the same way.

72%

of knowledge workers report less mental effort — and less critical thinking — when they lean on generative AI.

Microsoft & Carnegie Mellon

EEG shows AI-assisted writers form weaker memories and weaker brain connectivity — and recall their own work worse.

MIT Media Lab
17–25

Frequent AI use links to weaker critical thinking via cognitive offloading — strongest in younger adults.

Gerlich, Societies

These are associations, not a verdict. Walni's response is simple: the more you delegate to machines, the more it pays to keep your own attention, memory and reasoning measurably trained.

How scoring works

  • Each task stores raw per-trial data (accuracy and reaction times).
  • Normalized 0–100 scores are computed by deterministic, reproducible functions.
  • Task scores roll up into domain scores, then an overall cognitive index.
  • Difficulty adapts via server-side rules based on your recent performance.

What we don't claim

Walni is not a medical device and does not diagnose conditions. Scores reflect performance on specific tasks under specific conditions. We keep our claims within those boundaries.

Science & how it works · Walni