Skip to main content

The Lotti mental model

The useful friction is the point

Your life is not a backlog.

Lotti is a record of intent, decisions, reality, and learning—not one hungry list that flattens them all into “done” or “not done.” The plan is allowed to change. The record is allowed to tell the truth.

The loop: say it, choose it, live it, learn from it

01Give it a name

A category creates a useful boundary. A task names an outcome. Notes, recordings, images, and checklists keep the surrounding truth close.

Category · task · entry
02Choose today on purpose

Daily OS turns the work that matters into a proposed block. It is a decision for this day, not a rewrite of the task or a promise to a future self.

Plan · planned block
03Let reality answer back

Track the time you actually spent. Add the note, photo, measurement, or voice recording that explains what really happened—messy, useful, and intact.

Time record · evidence
04Notice the pattern

Habits, measurements, dashboards, and time analysis turn many ordinary days into something you can learn from. Then the next decision gets wiser.

Habit · measurable · dashboard
All screenshots
A Daily OS block says what you chose for today. The linked task and the time record remain their own honest things.
ScreenshotMobile view
Lotti Daily OS agenda with Project Waddle task thumbnails, planned blocks, and time context

The building blocks, without pretending they are all the same

Building blockUse it forExample
CategoryA meaningful area of life or workHealth, Family, Research
TaskAn outcome that needs progressPrepare the V1 release
ChecklistConcrete steps inside a taskVerify release notes
EntryContext or evidenceNote, image, audio recording
Planned blockA proposed slot in the dayRelease review, 10:00–11:00
Time recordWhat actually happenedRelease review, 10:18–11:07
HabitA recurring behavior you evaluateTake medication
MeasurableA numeric observationResting heart rate
DashboardA view over useful measurementsTraining overview

Agents are guests, not co-owners

An agent can read the context you give it, form an opinion, summarise a messy task, and propose a useful next change. Its report is an opinion with provenance, not a new fact in your history. Proposed task, checklist, status, or date changes wait for you to confirm or dismiss them. The narrow exceptions are an initial title or language for an otherwise empty task—after that, the same human-approval rule applies.

This is not politeness bolted onto a prompt. Lotti keeps the personal logbook and agent working memory separate, so the agent cannot treat your history as a scratchpad. You can use AI, tune it, or leave it entirely alone.

A practical first orbit

  1. Create categories that you genuinely use to make decisions.
  2. Capture work as tasks and keep supporting material attached to them.
  3. Use Daily OS to decide what belongs in a particular day.
  4. Record actual time and observations without forcing them to match the plan.
  5. Review habits and dashboards for patterns rather than individual perfect days.

Keep ownership clear

  • A linked task owns its title and category.
  • A Daily OS block owns its placement and duration.
  • A time entry owns the record of what happened.
  • AI proposes; you review and decide.
The point is not an immaculate timeline

If the plan changed, let it change. If the day went sideways, record what did happen. A useful system helps you see the shape of your life; it does not ask you to perform a more productive one for the database.

Next, create a first task or jump directly to Daily OS.