Product 02 / 07VOICE & INTERACTIONBeta · Apache-2.0 · separate asset terms

AI Meeting

Think out loud. Keep the next step.

Talk with an AI character, change direction, and keep the tasks that matter. Uncertain changes stay pending until they are confirmed.

Watch the film Discuss this technology

What it does

Conversation
confirmation
tasks

(03) AI Meeting — Film

(+) In detail

02 / VOICE & INTERACTION / Beta

Think out loud with an AI character

Talk it through. Change your mind.

You don’t need a finished plan to start talking. Work through what’s next with an AI character, say when your plans change, and save what you decide as tasks.

Try text tasks — no sign-up Read the voice-access conditions

The first button opens the account-free task UI in Japanese. Voice is a separate verified-email, time-limited entry point.

Try the guided sample first Sample only. No AI run or external action. Editing needs JavaScript.

FIRST THING TO TRYAdd one task, then try completing or deferring it in the text UI.

Setup, scope & license

Apache-2.0 · separate asset terms · Account-free tasks and the verified-email, time-limited voice trial are different experiences. External meeting connections require setup and evaluation.

No sign-upAdd, complete and defer tasks in the browser
Quoted, not guessedTasks and due dates come from the spoken words and wait for review
JSON exportTake saved tasks out as JSON; nothing syncs automatically

WHAT THE FILM SHOWS

“Actually, make that tomorrow.”

Speak or type
Capture work and optional deadlines in a conversation or text.
Review the change
Potentially misheard changes remain distinct from saved tasks until confirmed.
Keep the next step
Return to the saved task state instead of losing the outcome of the conversation.

TRY IT

Start with one thing to do tomorrow.

Add, complete or defer a task in the Japanese text UI without signing up. To try voice, first read the separate verified-email and time-limit conditions.

Try text tasks — no sign-up Inspect on GitHub

CURRENT BOUNDARIES

Availability and development.

Current scope

Account-free tasks and the verified-email, time-limited voice trial are different experiences. External meeting connections require setup and evaluation.

Apache-2.0 · separate asset terms

Read the evidence

Proposed interface design

The separate design film is a proposal for a future interface. It is not a demonstration of the current product or proof that the proposed UI has shipped.

Open the labelled design study

(05) What it can do

An interruptible conversation that leaves confirmed tasks.

Source: Check the README · checked against the repository on 2026-09-27 (3b269e9)

F-01 / TALK

Talk it through with an AI character

A one-to-one voice conversation you can redirect part-way. Voice is a separate verified-email, time-limited entry point.

Voice conversation
Interruptible
Gated entry

F-02 / TASKS

Tasks from quoted speech

Task titles and due dates reuse the words that were said; owners or deadlines that were not spoken are never invented. Possibly misheard changes stay pending until confirmed.

Quote-based
Confirm before save
No assignee field

F-03 / KEEP

Pick up where you left off

Confirmed tasks stay in this browser and can be exported as JSON. There is no automatic sync to meeting tools or external services.

Browser storage
JSON export
No auto-sync

(06) Scope and evidence

What it cannot do,
stated up front.

Public sources checked 2026-09-15. Check each repository and product page for current terms.

SCOPE

Account-free tasks and the verified-email, time-limited voice trial are different experiences. External meeting connections require setup and evaluation.

PROOF / PUBLISHED EVIDENCE

45-second recording with real Gemini responses, synthetic speech input, and some scripted confirmations.

(07) Build with it

Prototype voice-based training, task capture, or an AI character experience.

Discuss a project