The finder has evolved

What happens to branding when people stop searching and start asking AI agents to find, compare and recommend on their behalf? Discovery becomes an interpretation problem: an agent reads the website, extracts a few passages and presents its own synthesis. The question shifts from what should people remember about us to what must a machine understand to represent us accurately. Brands need to be easy to love and difficult to misrepresent.

The framework behind the talk had three layers:

• Context: structured access to a brand's identity and rules.
• Insights: how the brand appears across every source.
• Actions: agents that turn those signals into work, with human oversight.

Together, these layers presented brand systems as active infrastructure rather than static guidelines. The aim was not to automate every creative decision. It was to make reliable brand intelligence available wherever work happens while reserving expert attention for the decisions that require it.

It closed by swapping the familiar question, will AI take design jobs, for a more useful one: is your taste worth citing?