We research what AI can already do in practice: which work it can take over, and how much faster and better that work becomes. And we study what that does to people: to jobs, purpose and meaning.
Bridge is our own test bed. Everything we make is built with AI agents, so we see first-hand which tasks they take over, where they make work faster, and where they still fall short.
designer behind Bridge, working with AI agents
months from the first line of code to a full creative suite
lines of code
of all commits co-written by AI agents
Source: Bridge's own code repository, October 2026.
Design, video, animation, audio and 3D. How far can agents get from a description to a finished piece, and where does a human eye still decide?
Writing, research, presentations and bookkeeping. Which tasks disappear, which ones change shape, and how much time does it really save?
Building and running tools with agents. What happens when anyone can make the software they need, and let agents operate it?
The faster AI takes over work, the more urgent the human questions become. These are the six we work on, in the open, with anyone whose work is changing.
A job is a bundle of tasks. Agents are taking that bundle apart, one task at a time. What remains when the routine is gone, and is it enough to fill a working week, or to pay for one?
When software, design and text can be made on demand, the price of making collapses. Millions of people earn their income from exactly that. Where does their work go, and who captures the value instead?
For many people, being good at something is part of who they are. When that skill stops being scarce, what is left? Craft for its own sake, care for each other, or something we have no word for yet?
Technology now moves faster than schools, laws and careers can adapt. Faster is not automatically better. When is it wise to slow down, and who gets to decide?
Building your own software feels like independence. But your agents still run on someone else's models, servers and data. Freedom for individuals could quietly turn into dependence on a handful of companies.
Not how to use a specific program. Maybe how to ask good questions, how to judge what a machine makes, how to work with other people, and how to know what you actually want.
We build the AI tools we study. Bridge Suite lets us, and you, experience this future first-hand instead of reading about it.
We track what agents actually do, task by task: what they take over, how much time it saves, and where people stay essential.
Essays, field notes and conversations with people whose work is changing: makers, teachers, entrepreneurs, unions and policymakers.
Name your profession. See what may shift to agents, what stays human, and a question worth asking yourself.
Tell us what is changing in your work. The stories we receive shape what we research and write next.