Idea
Will the most useful AI agent be the one we trust enough to let in?
As agents become capable enough to act, their practical usefulness depends on context, permissions and delegated authority. Trust is part of the architecture: identity, boundaries, auditability, approvals and revocation determine what an organisation is willing to let an agent do.
Current thinking
The best AI agent may not be the smartest one. It may be the one you trust enough to let in.
Explore access to email, calendar, files and organisational systems; distinguish read access from authority to act. Build from a business problem rather than searching for an AI use case. Governance should enable graduated, accountable delegation, not just impose a blanket restriction.
Writing on this idea
-
Note
The Scarce Resource May Be Trusted Access
As agents become capable enough to act, the limiting factor may stop being what the model can do and become how much context, permission and authority we are willing to give it.
-
Note
The Problem With a Blank Prompt Box
We've given people extraordinarily capable AI tools and often left them to imagine the use cases. Perhaps discovering what is possible should be part of the product.