Jonathan Grizou · We are our choices

This interface can read your mind

Last update: August 2019

A web-app to experience the power of self-calibrating interfaces.

Quick access: [Crack it] [Try it] [Grasp it]


Below is a typical user interface, it is a code entering device. With two big colored buttons, it is designed to remove most user choices and forces you to behave in one standardized way.

The user clicks on two colored buttons.

The aim is normative. It is more efficient and convenient if we all behave the same way. And with apps like Instagram or Twitter, designers have done a great job at funneling our interaction.

But let me show you another way to interact with digital devices.

Below is the same interface but buttons have no predefined colors. You get to decide the buttons’ color in your mind, and the machine adapts to your preferences on the fly.

User 1 choice.
User 2 choice.
User 3 choice.

I call these “self-calibrating interfaces” because they adapt to your preferences while you are using them.

You will find below several options to play with theses interfaces and gain a deeper understanding of how they work.

Learning tracks

Quick access: [Crack it] [Try it] [Grasp it]

Crack it - the open vault challenge

When you teach a child something you take away forever his chance of discovering it for himself - Jean Piaget

If you want to really get it, you should re-invent it for yourself. So I have designed a challenge tailored for participants to gain an intuitive understanding of self-calibrating interfaces.

In this challenge, you will see me entering a secret code in various interfaces. Your goal is to crack the code and enter it correctly into the interface. By doing so you are forced to think like the machine.

The challenge is divided into levels of increasing complexity, designed for a progressive learning experience. Clues and take home messages are available if you get stuck, but I encourage you not to use them.

You are the master of your education. If you do the challenge, don't cheat your way through, struggling is part of the learning process.

Go to challenge

Try it

If you do not have time for the challenge, you can directly try any version below. If this is your first time, I recommend going through them in order.

For each version, a small explanatory text is provided along with a didactic video.

Btw, the audio version can be especially entertaining as you can replay previous recordings in ludicrous ways.

Go to demo

Grasp it - look under the hood

If you want to see how it works, I have built a little window into the machinery powering the interface. It provides detailed feedback about the current state of the machine.

It allows you to challenge the interface in complex ways. For example, I sometime use these advanced versions to fool the machine into making a false prediction, or to select my actions such that it takes the longest possible time to identify a digit.

Go to tutorial

Wait but why?

Because I have consistently failed to explain what a self-calibrating interface does to other researchers, friends or family in a few minutes. This web application is my attempt at explaining, showing, and getting people to feel what a self-calibrating interface does in one minute on my phone.

I also have the intuition that this technology can be useful. But I cannot figure out how on my own, I am too involved in the research and find it hard to think outside the box. By implementing this technology into an easy to share demonstration, I am hoping to reach a variety of people. And maybe one day someone will figure out a place to apply it for good. If you are that person, feel free to contact me.


A 2-page paper is online and provides more background.

Detailed information, papers and code available on this project page.

I have more to say than time to write, so feel free to get in touch.