AR Glass + Face Lookup

On a live stream, I looked at a colleague through a pair of smart glasses, and the details they had chosen to share appeared on my screen while we talked.

The challenge

We'd seen an online demo of the dark side of this technology: smart glasses plus public face search, used to look up information about a stranger on the spot. We wanted to show both sides. Yes, your face can be searched. But with consent, the same idea can help: at a company event, a speaker could see what the people in the room care about and shape the talk around them.

Consent was the design constraint. Only colleagues who opted in were in the system, and the only information it could return was what they had chosen to share, such as their areas of interest or expertise.

What I built

The glasses stream what the wearer sees. When a face in view matches someone who opted in, a web page beside the speaker shows that person's shared details, so the speaker has context about the room in real time.

How it came together

The streaming workaround added a few seconds of delay but proved reliable, and matching was very accurate during our live-stream demos. Nothing broke. The limit was the hardware: today's glasses have no real display and no SDK, so the results lived on a separate screen.

What happened

We demonstrated it on a live stream. Colleagues who had opted in saw their own shared profiles appear as I talked to them, and were impressed. Others were uneasy, which was part of the point: the demo made both sides of the technology concrete.

What I'd take forward

I'd design it around play rather than people. A kiosk where guests opt in, take a photo and answer a few questions they're happy to share works for a conference; better still is a game, where what the system recognizes you by is what you did in the experience, not who you are.