Future Interfaces.
How humans and AI systems negotiate work, intent, and judgement. The disappearance of interfaces, the emergence of conversational and agentic abstractions.
The interface has been shrinking for forty years — command line to window to touch surface to text field to, increasingly, nothing the user consciously operates. Each step moved work from the person to the system, and each step changed who was able to participate.
The current step differs in kind rather than degree. When the interface is a conversation and the system can act, the design question stops being how a user expresses intent and becomes how much judgement the system is permitted to exercise on their behalf — and how that permission is granted, bounded, revoked, and audited afterwards.
That is not a UI problem. It is a delegation problem wearing a UI. This territory is an intentional reservation: the questions are sharp, the published analysis is not here yet.
What this territory tracks
- Delegation: what a system may decide without returning to ask
- Legibility: how an agent’s reasoning is made inspectable after the fact
- Recourse when an autonomous action turns out to be wrong
- What else disappears when the interface does
A territory the firm watches and writes about, but where no published artifact has landed yet. The shape of the inquiry is here; the analysis is in progress.
Subscribe to the feed to see first publications when they arrive.