From response to initiative.
Systems that can pursue goals over time change the meaning of a tool. The central design question becomes permission: what may act, for whom, and under whose review?
Future intelligence observatory
The futureis learningto look back.
We explore the intelligence emerging between people and machines—and the futures taking shape around it.
Six lenses
Watch the relationships, not only the models.
The future of AI will be shaped as much by the roles we give it as by the capabilities it gains. These are the fields we watch.
When making becomes choosing.
When possible drafts become abundant, authorship moves toward direction, taste, and judgment. The work is not merely to generate more. It is to decide what deserves to exist.
Discovery at a different tempo.
AI can help compare more possibilities, surface unexpected connections, and shorten some paths from question to experiment. It does not replace evidence; it raises the value of asking better questions.
Capability enters the physical world.
Once intelligence can sense, move, and intervene, mistakes carry weight. Restraint, legibility, and graceful failure become part of what competence means.
Intelligence becomes infrastructure.
As AI mediates decisions across organizations and public life, its assumptions can become policy in practice. The future depends on who can inspect, contest, and change those assumptions.
Humanity after exclusivity.
We have often defined ourselves by abilities we believed were ours alone. As that boundary moves, meaning may shift from what only humans can do to what humans choose to value.
OBSERVATION · 01
Intelligence is leaving the interface. It is becoming memory, agency, infrastructure, collaborator, and environment.
The question is no longer whether AI enters the future. It is what kind of future enters through AI.
Three horizons
Not predictions. Positions from which to look.
Each horizon is a scenario—a way to examine today's choices from a different distance.
01NOW
AI beside us.
Intelligence appears through familiar surfaces: the document, the workshop, the studio, and the lab. The immediate work is learning where assistance sharpens judgment—and where it quietly replaces it.
02NEAR
AI between us.
Agents may begin coordinating across people, software, and institutions. The decisive questions will concern delegation, accountability, memory, and the right to interrupt.
03DEEP
AI around us.
Intelligence may recede from view and become part of the environment: ambient, adaptive, and woven into daily systems. At that scale, values cannot be added later. They become architecture.
A scenario is not a promise. Its value lies in what it reveals about the present.
A human position
The future should remain answerable to life.
We are not here to worship intelligence or fear it. We are here to understand the relationships it makes possible—and the power those relationships carry.
A worthwhile future does more than automate what exists. It enlarges human agency, protects room for refusal, and gives more people the means to understand, create, care, and participate.
Capability deserves wonder. Consequence deserves equal attention. We choose curiosity without surrender, ambition without inevitability, and progress that remains open to question.
The point is not to predict the future. It is to become more responsible for the one we are making.