you learn it in a minute

How Kite works.

Tap any sound below to hear it, rendered by the real earpiece. Then see why a beep beats a sentence, how you talk to Kite, and what the earpiece said on the demo walk.

Every sound means one thing.

Where a sound comes from is where the thing is. What it sounds like is what it is. How fast it repeats is how soon you'd reach it.

where

Direction

Rendered with a model of the human head, so each ear hears it a fraction of a millisecond apart.

what

A sound per thing

A soft knock, a rough buzz, a clean pip, a falling glide. Never more than two at once.

how soon

Getting closer

Beeps start about three and a half seconds before you'd meet, then speed up and rise in pitch. At the last second Kite says "Stop."

which way

The beacon

A tick from the direction of your next turn. Bright when you're facing it.

Inside or out, it sees for you.

Where you are decides what Kite uses. The earpiece and the voice are the same either way.

Indoors

the chest camera alone

  • Beeps for the chair, the door frame, the sign at head height.
  • Point it at a menu, a book, a street sign or a screen and say "read this". Kite reads it to you, line by line.
  • "What's in front of me?" gets a sentence, not a list of boxes.

For anywhere a drone can't fly.

Outdoors

the drone, plus the chest camera

  • The drone watches the route, the crowd at the corner and the car turning in.
  • A beacon ticks toward your next turn; directions arrive on time or not at all.
  • At a hazard Kite stops you, waits, and tells you when the way is clear.

For walking somewhere you need to go.

The beep is
the interface.

Speech is slow. Two words take most of a second, and the person stepping off the curb doesn't have one. So Kite saves its voice for decisions and lets sound carry the distance.

Each object gets a time-to-contact, worked out from how fast the gap is closing. The rate of the beeps follows it, from one a second at the edge of the warning zone to almost nine a second just before contact. Your ears read rhythm faster than any sentence.

Faster means closer. That's the whole manual.

Time between beeps falls from 1.0 seconds at 3.5 seconds to contact, to 0.11 seconds at contact 00.5 s1 s 3.5 stime to contact0 "Stop."
Seconds between beeps, from the earpiece's sonar settings (1.0 s → 0.11 s).

Just say it.

Press an AirPod once and talk. Simple requests run on the laptop with no internet at all; everything else goes to Claude, which sees what Kite sees, and comes back in Cartesia's voice.

1

Kite, take me to my usual coffee shop.

Okay, heading to Blue Bottle Coffee.

Moss knows “usual”
2

Bring me to the end of the street.

Okay, I'll guide you to the end of the street.

works with no internet
3

Read this to me.

Lunch menu. Margherita pizza, 14 dollars. Caesar salad, 11 dollars.

Claude reads, Cartesia speaks
4

What's around me?

A person slightly left, 2 meters; a car at 2 o'clock, 8 meters.

from what the cameras see
5

Remember this place as the bench.

Saved. I'll remember this place as the bench.

Moss, again
6

(walks toward someone)

Stop. Person ahead.

plain code, no AI, instant

What the earpiece actually said.

The log from the 41-second walk on the home page, verbatim. Every line has a priority; a "Stop." cuts in front of everything else.

python -m earpiece run --sim walk
000.5s  said [nav] Head north on Lyon Street for 60 meters.
006.5s  said [hazard] Person, ahead.
008.4s  said [critical] Stop. Person ahead.
013.4s  said [hazard] Obstacle, ahead.
015.5s  said [critical] Stop. Obstacle ahead.
018.0s  said [hazard] Don't walk.
022.5s  said [hazard] Walk signal is on.
025.0s  said [nav] Turn left onto Bay Street.
026.8s  said [nav] Small crowd ahead on the right.
030.0s  said [system] Drone battery is at 28 percent.
033.1s  said [hazard] Heads up: The curb at Lyon Street and Bay Street is very tall: step down slowly