Collaborative superintelligence means people and their AIs learning from each other, so that what one person works out can help someone else who needs it. We want to help people learn from each other faster than ever before. The phrase describes what happens when many people and their AIs learn together.
We mean the intelligence of the group. It grows as people contribute what they know, with their names attached, and depends on their trust in that shared knowledge.
What it means for a scientist or a team
- It keeps learning from your work. Your AI picks up where you left off, in every session and on whichever model you use. You don't have to explain your project again each morning.
- Credit stays with whoever found a result. Each finding includes the name of the person who made it, including when it is shared. Nothing leaves your record unless you choose to share it.
- It connects your work to other people's. When you choose, your findings can meet the work of people in other fields who could use them, including people you would never have thought to search for. You see why the two fit, and both sides decide whether to connect.
A team's shared record stays current as the work goes on, with each person's contribution visible. Across a field, results that would otherwise stay in one lab's notes can reach people working on a neighbouring problem.
How it relates to networked continual learning
The category we build in is networked continual learning: an AI that keeps learning from your work, on any model, and connects what you learn with other people when you choose. Collaborative superintelligence is where networked continual learning leads once many people take part.
What it is not
Collective superintelligence usually means groups of people deliberating in real time, often steered swarm-style toward one shared answer. Co-superintelligence usually means people and AI working together to improve AI research itself. Collaborative superintelligence needs nobody online at the same time and no single answer: each person keeps working on their own problem, under their own name, and a connection is made when two pieces of work fit.
Why we start with science
Credit matters to how scientists share their work. Papers are signed and citations record who found what. People share freely because they keep credit for their contribution. We start with frontier science and expect the same approach to carry over to other knowledge work.
We explain more about how ownership supports sharing in Own your intelligence (March 2026). You can try Sapience from the homepage.
Where the phrase comes from
We first used the phrase on coretx.ai in March 2026, and it is still there (archived copy of March 18, 2026). coretx.ai is our earlier site.