John Rice — PlaygroundA prediction, recorded before we know the answer.
September 9, 2026

Future Hardware Prediction

The next personal computer may not be a screen in your pocket. It may be something you wear: an intelligent earpiece and a ring.

A hand wearing a minimal ring points into an empty field of light.

The prediction

AI has made an old contradiction obvious. We can ask an agent to reason through extraordinarily complicated work in ordinary language, while our phones still make us navigate screens, apps and settings to perform simple actions.

The next leap is not necessarily a better phone. It is removing the need to take the phone out.

Ear + Ring + Agent

The earpiece is the private conversational channel. The ring supplies a deliberate hardware affordance: invoke, dismiss, confirm, interrupt—and point at the physical world. The agent supplies the intelligence.

A person wearing an earpiece and ring points into the physical world.

Why a ring?

A real button creates an unambiguous boundary around agency. You know when the system is listening because you intentionally invoked it. The hand also provides something language needs constantly: reference.

Point at a lamp and say, “Turn that off.” Point at a building and ask, “What is that?” Point at a display and say, “Put it over there.” The physical gesture establishes this, that and there; language carries everything else.

IR, inertial sensing, UWB, cameras or other spatial systems can fill in the implementation. The enduring interaction is simpler than the technology underneath it.

The interaction grammar
Point.
Click.
Talk.
Talk. Click.
A ring-wearing hand makes a deliberate pointing gesture.

The hardware interface can remain simple even as the intelligence becomes dramatically more capable. New capabilities do not require new buttons, menus or apps.

Culture, not gadgetry

The form matters. Earrings, ear cuffs and rings already have cultural permission. They can be stylish, personal, collectible and visible—metal, ceramic, jewelry, fashion collaborations, different forms for different people.

That may be crucial to adoption. The product does not have to look like a computer strapped to your body. It can be something people already understand how to wear and show off.

The phone can remain in the pocket initially as battery, radio, compute, identity and fallback display. Screens do not disappear; they become resources the agent summons when information actually needs to be visual.

A ring-wearing hand points toward a simple physical object.

The test

Battery life and cultural adoption are the immediate constraints. But if the system becomes reliable, the old behavior could invert surprisingly quickly.

Today it is normal to take out a phone, unlock it, open Maps and check where to go. In a successful post-phone interface, that may eventually look as cumbersome as unfolding a paper map does now.

This page is being published on September 9, 2026 as a timestamped prediction. We can come back later and see how close—or how wrong—it was.

Questions that follow

The hardware is only the beginning. If an agent becomes something we wear, consult continuously and eventually think alongside, these may be the more consequential questions.
01
If an AI is the private voice in your ear for decades, who should legally own the model it builds of you?
02
Does a personal agent need a protected relationship with you—something closer to attorney-client privilege than a software privacy policy?
03
When your immediate desire conflicts with goals you established over years, which version of you should the agent serve?
04
If an agent sees that you are about to seriously harm yourself, when does assistance become a duty to intervene?
05
If children grow up thinking alongside an agent, does it strengthen the inner voice—or gradually replace part of how that voice develops?
06
When the useful unit becomes human + agent, where does human agency actually begin and end?
07
Can an agent truly represent you if advertisers, governments or its manufacturer can influence what it recommends?
08
Should everyone have a baseline private agent, or will subscription tiers create different classes of cognitively augmented people?
09
Could an agent preserve authenticated evidence of your intent—medical wishes, contracts, a will—and continue representing those wishes after you can no longer speak?
10
When you die, who inherits your agent? Should it preserve part of you, serve your estate—or have the right and obligation to disappear with you?
11
If your agent can learn and improve itself, how can you know that the agent you trusted yesterday still serves the same interests tomorrow?

A coda

And perhaps that last question isn't really about AI.

How can you know that the person you trusted yesterday still serves the same interests tomorrow?

Humans learn, adapt, conceal, change their minds and become different through experience too. Trust has never required another mind to remain unchanged. It requires ways of living with the fact that it can change.