The ordinary morning

Noor walks past a Georgian bakery she has never been to. Camera up, one tap — a menu appears, ordering is enabled, nearest collection is 7 minutes. She walks home with bread and has not typed a word.

Repair becomes a photo

Her boiler is leaking. She does not remember the model, cannot describe the fault, and cannot find the manual. Camera up — the model, age, and fault are recognised; a GeraHome quote for a plumber appears; a booking is made for Thursday morning. No phone call, no form, no search for the right term.

Insurance becomes visual

Her car has a scratch. Camera up — the scratch is assessed, repair options priced, GeraSure notified, and a renewal quote adjusted. The “claim form” is a photograph plus a confirmation.

Small businesses become discoverable

The Tbilisi plumber from the previous posts has a storefront. In 2030 the storefront is machine-readable — a QR-adjacent signal, a published descriptor, or just a clear sign — and his booking endpoint routes inbound camera actions. He competes on availability and price, not ad spend.

The lines that do not move

Some things stay off-limits permanently. Face recognition of strangers. Identification of minors for commerce. Private- environment detection (hospital beds, bedrooms, bathrooms). Automated assessment of distressed situations without a human in the loop. These are bright lines — not because they are infeasible but because the downside cost is too high and the upside benefit is not large enough to justify building past them.

Public spaces get a soft layer

Museums, transit stations, airports begin to publish machine-readable visual descriptors — “this sign, this exhibit, this gate” — so agents can read them without scraping. This is the physical-world version of MCP: public venues declare what their visual surface means.

The risks

Surveillance creep. Every visual-intent query is a camera event. The audit layer, refusal policies, and on-device gate are what keep the product safe.
Cultural visual vocabulary drift. What is a “storefront” in one country is not in another; mislabelling produces wrong actions.
Accessibility. Not everyone can hold a camera. The voice- first alternative via GeraVoice is a required complement, not optional.

The ordinary shift

Most of the 2030 impact is boring: less typing, less searching, faster local discovery, more completed transactions for small businesses with good physical presence. Typed search still exists. The camera just becomes the first finger the user reaches with.