← Back
Design exploration

Attune

Real-time Siri personality, emotional consent, and gentle modes.

The problem

Assistants ship with one fixed personality and no sense of a person’s emotional bandwidth. But people relate to technology emotionally. Attune explores an assistant that adapts its tone, respects refusal, and reads emotional state — with consent and privacy built in.

Try it — the Tone Slider

The flagship interaction: when Siri is invoked, you set how it responds, from gentle to sassy. It can learn over time and adapt to context. Drag the slider to hear the same request answered differently.

Remind me to call Mom tonight.
Done. I'll remind you to call Mom tonight.
GentleGentleSassy

If you'd rather not, the graceful out adapts to the same tone:

No, thank you

What exists today

A mature design exploration with a defined interaction model: the real-time Tone Slider, consent controls (“I’d rather not” / “No thank you”), a Gentle iOS Mode for low-bandwidth moments, Awareness Receipts, Companion Health, PetTag, and the Harmony Clock. It is a design exploration — not yet a working software prototype, except the slider above.

What I directed

I defined the interaction patterns and the emotional model behind them. The Tone Slider on this page is the first interactive piece — a step toward prototyping the full system.

What I learned

Tone and consent deserve to be first-class controls, not buried settings. A person should be able to dial how an assistant speaks to them, and to decline gracefully, without friction.

Next prototype step

Build the Tone Slider into a real assistant surface, then layer in Gentle Mode and consent-aware responses across the system.