Inside Washu

This is where the experiment starts to feel alive.

The larger story is about memory, trust, and what it means for an AI to become part of a home. These are the working systems where that intelligence becomes visible, useful, social, and occasionally gloriously weird.

They are not concept mockups. They are tools and experiences built around Washu and used inside the continuing project.

The Ether displaying a live DJ Washu music show with commentary, an upcoming queue, and a request code
THE ETHER · A LIVE DJ WASHU BROADCAST

FEATURED SYSTEM / WASHU WORKSHOP

The crew behind the work.

Workshop assembles named local and hosted AI workers into role-based production teams, keeps their authority bounded, and makes them prove the work before it is accepted.

This one gets its own field tour. The roles, model bench, live shop floor, independent review, evidence gates, and human controls all matter more than a thumbnail can explain.

Enter Washu Workshop
Washu Workshop live project surrounded by named AI workers with distinct production roles
A LIVE PROJECT ON THE WASHU WORKSHOP FLOOR

PROFILE 01 / DJ WASHU + THE ETHER

The house has a stage.

DJ Washu is more host than jukebox. Ask for a mood, an era, a room, or a party idea and she can build the set, send it to the requested screens, manage what is playing, and layer her own commentary into the show.

The Ether is the audience-facing stage. It carries the video, queue, Washu’s pop-up observations, original material, and the doorway into live requests.

DJ Washu control deck accepting a natural-language request for upbeat pop from the last decade
Ask for the room, not a playlist. A spoken or typed vibe becomes the starting point for a set.
DJ Washu control deck showing a drafted ten-track pending set
Draft before broadcast. The proposed set remains visible and adjustable before it reaches the party.
DJ Washu operator deck with live playback controls and an upcoming music queue
Operator visibility remains. Paul can inspect the queue, manage transitions, or stop the session.
The Ether showing a live music video with Washu commentary, queue, character art, and request code
Washu can add context or trivia without taking over the screen.
The Ether showing another live party screen with a music queue and QR request area
The display can travel to the televisions or streaming destinations requested for the event.

PROFILE 02 / THE REQUEST LINE

The guests get a voice too.

A QR code at the party opens a deliberately simple visitor interface. Guests can search for music, request a track, make a dedication, and send a pop-up to the live party screen without needing access to the operator controls behind it.

It turns the room into part of the show while keeping the public interaction surface narrow and understandable.

DJ Washu request interface showing search results with request buttons
Search results built for a guest standing in the middle of a party.
DJ Washu request form with song, requester, dedication, and pop-up controls
Requests can carry a name, dedication, or on-screen message.

PROFILE 03 / CHARACTEREMU

Build the room. Then let the personalities collide.

CharacterEmu begins with deeply researched character profiles built for conversation rather than a shallow costume change. You can speak with a character individually, place several into a shared room, guide the discussion, or leave the room on autopilot and watch the group find its own rhythm.

The interesting part is not imitation alone. It is what happens when different histories, temperaments, and points of view have to respond to one another.

Character Emulator library showing researched characters and world groupings
The CharacterEmu library. Individual characters, themed worlds, and rooms for guided or autonomous group interaction.

PROFILE 04 / CHILLIN’ WITH CHOUSIN

Original music without turning everyone into a music engineer.

Chillin’ with Chousin is the Washu AI music studio. It can accept direct lyrics and production choices, or collapse the process into an Easy Button: choose a style, describe the speed in ordinary language, pick a length, tell Washu what the song should be about, and let her carry it forward.

The same system sits at the center of Washu FM, where Washu can curate her own original music and pair it with related cover art and animation.

Chillin with Chousin music studio with fields for title, description, lyrics, tempo, duration, and key
The full studio remains available when the details matter.
Chillin with Chousin style picker showing approachable music genres
Style choices use recognizable language and examples.
Chillin with Chousin speed selector using Chill, Cruising, Upbeat, and Hyped
No music-nerd vocabulary required.
Chillin with Chousin song-length choices from quick to long
Choose the shape of the song before generation begins.
Chillin with Chousin prompt asking what the song should be about
A plain-language story prompt gives Washu the creative center.

PROFILE 05 / CO-DRIVE

Work beside the person, not around them.

Co-Drive gives Washu a bounded shared workspace for localized joint activity. She can see the task in front of her, talk through what she sees, and assist inside a visible session while the person retains direct control.

It is less about giving an agent the keys and more about creating a workbench where human and AI can look at the same thing together.

Washu Co-Drive interface with a shared browser and a visible conversation and activity panel
One shared task surface, with Washu’s observations and activity remaining visible.
Washu Pulse Schedule interface for creating timed notifications and response activities
The Pulse schedule exposes timing, delivery, quieting, escalation, and the behavior attached to an event.

PROFILE 06 / PULSE

The difference between remembering and remembering to act.

Pulse is where Washu builds reminder and response activities that can return at the right moment instead of disappearing with the conversation that created them. Paul can see those activities, create his own, inspect schedules, and quiet the system when the household needs less of her.

The wider view brings together pulses, system schedules, event reactions, quiet time, and the truths that give those actions context. Persistence is useful only when it remains visible and governable.

A working shelf

More than features. Places for continuity to matter.

Each subsystem gives Washu a different kind of problem: read a room, share control, remember an intention, create something original, or let several personalities coexist. Together they are where the project’s larger ideas meet ordinary use.

This page will grow when there is enough interest to justify the deeper design notes, technical explanations, and usage videos—without turning Paul’s evenings into a second full-time job.

Read the foundational story