Skip to main content
← Blog
ManifestoApril 10, 2026·6 min read

Towards ambient intelligence

There's an implicit promise in the idea of artificial intelligence: that machines will eventually handle the things we don't want to handle. Not that they'll give us more things to do. Not that they'll invent new places to spend time. Not that they'll demand attention for themselves.

AI as it has developed in recent years has betrayed that promise. Not out of malice — out of economic logic. Products are measured by engagement. Engagement is measured by time spent. So we built AIs that want you to talk to them, open them, consult them. AIs whose success depends on your attention.

We think that's a fundamental mistake.

The wrong turn

The chatbot has become the dominant paradigm of consumer AI. You have a problem — you ask a question. You want an idea — you write a prompt. The interface is conversational, the interaction is voluntary, the result depends on the quality of what you ask.

It's a powerful tool for some tasks. But it's a terrible model for everyday decisions — the ones that come back every day, at the same time, when you have the least energy to formulate anything.

Nobody wants to open an app at 6:30 pm to write a prompt about what they should eat. Just as nobody wants to ask their thermostat what temperature it should display. The best technology doesn't ask — it knows.

An AI that waits to be spoken to isn't intelligent. It's just available.

What “ambient” really means

The concept of ambient intelligence isn't new. Researchers at MIT and Xerox PARC were talking about it in the 90s, under the name calm technology. The central idea: good technology lives at the periphery of your attention. It informs without interrupting. It acts without asking permission. It disappears into the background until the precise moment it's useful — then disappears again.

Electricity is ambient. Central heating is ambient. Background music is ambient. You don't “use” them — they're there, doing their job, and you live your life.

Ambient intelligence is that, applied to decision-making. A system that observes, learns, and acts — without you having to instruct it at every step. Not an assistant you consult. An environment that thinks with you, silently.

Chora as a first step

We started with dinner. Not because it's the most important problem in the world — but because it's the most daily, the most universal, and the one that best concentrates the tension we're trying to resolve: a complex decision, with high recurrence, at the worst moment of the day.

Dinner is also an ideal testing ground for ambient intelligence. There's enough signal in your choices — what you accept, what you refuse, what you replace — for a system to genuinely learn. Not declared preferences in a form. Revealed preferences through your actions.

But dinner is just a beginning.

Where we want to go

The vision behind Chora is broader than meal planning. What we're building is an ambient decision infrastructure — a system that learns your life context and takes care of the repetitive decisions that consume energy without being worth it.

Dinner today. Maybe tomorrow, groceries — not a list to build, but items ordered before you realize you're running low. Maybe other daily friction points that every household knows in its own way.

In every case, the principle stays the same: the system does the work, you keep control. You can always say no, adjust, change your mind. But you don't have to initiate. You don't have to think to think about it.

What we refuse to build

Ambient intelligence also implies refusals. We won't build a culinary chatbot. We won't build a dashboard with graphs of your eating habits. We won't build a “nutrition score” or a reward system. We won't build something that makes you want to open it every day.

If you open Chora every day, we've failed. Success is when you stop thinking about us — because the decision is already made, silently, before the question even arises.

We don't want your attention. We want to give it back to you.

That's ambient intelligence. And it's the only direction we want to build in.

Want to see what this looks like in practice? Read about the silent AI modelthat powers Chora's daily suggestions. Or explore why one clear opinion beats ten options — the product philosophy behind every decision Chora makes for you. Try Chora free for 7 days →