The curious are rewarded
The mark on this site is an attitude indicator. I'm a pilot, and it's the instrument I look at most. Here's what's going on in it, and why each part is there. The link at the end of each section takes you to the research it points at.
It shows an attitude
An attitude indicator draws the horizon the way the aircraft sees it: sky on top, ground below, and a fixed little aircraft symbol in the middle. Nose up -> horizon drops. Bank -> horizon tilts. It sits in the middle of every panel because it's the one you check most. The mark is frozen at 15 degrees nose-up, wings level, which is what it shows in the first few seconds after takeoff.
It's also the oldest answer to the question my thesis asks. Most interfaces assume the room holds still. A cockpit never does, so the instrument moves the world and keeps you fixed. My PhD is basically about that inversion: what happens to technology when the room moves (a van), shrinks (a small apartment), or isn't really a room at all (a card table).
It's always day zero
Pilots call the moment the wheels leave the ground "rotation". At that point nothing is proven yet, you have the least altitude you'll have all flight, and the instrument shows you exactly what you're doing. I like that as a way to work. Every project on this site started at rotation: a question, a field site, and an artefact that didn't exist yet. Proteus was a lab prototype before it was a company. Bodhika was a study with 9 parents before it was a lab with a school in it. The climb is the part with the most decisions and the least margin, and it's the part I actually enjoy. Day zero, every day.
It's a hexagon because of a chart
Aircraft instruments are round, so why a hexagon? On aeronautical charts, a VOR (the ground beacon that aircraft navigate by) is drawn as a hexagon with a dot in the middle. You fly from one VOR to the next. Each one only tells you where the next one is, and that's been enough to cross continents for 70 years.
The research has gone the same way: a van -> small homes -> a card table -> a classroom in Hyderabad. I didn't plan that route. Each site was just the next beacon in range from the last one. The About page says "where good problems take me" for the same reason.
Two bars and a dot
On most attitude indicators the aircraft symbol is two short bars and a dot. Those are also the three shapes in the oldest experiment in HCI: two targets and a pointer moving between them. Paul Fitts ran it in 1954. The result (time to reach a target depends on how far away it is and how big it is) still decides how big the buttons on your phone are.
Fitts kept the targets fixed and measured the pointer. Proteus flips it - the target is programmable. A playing card that can become any other card, on a table where people have been reaching for cards for a few hundred years. The RQ: how much can you change about an object before people stop treating it as the thing they know?
The favicon turns left
In the browser tab the instrument is banked 20 degrees left, with the aircraft symbol removed. At 16 pixels, two bars and a dot stop looking like an aircraft and start looking like a face (it did, and it was bad). The left turn isn't random either - circuits at most airports are flown with left turns, so after takeoff, left is the way you go.
The face is the same problem the Bodhika study is about, at 16 pixels. Shrink something or pull it out of context, and people read it as something else. A wellbeing score on its own reads as a label on the child. The same score inside a whole picture of the child reads as information a parent can act on. Same reason the dashboard got redesigned.