What should
we expect from
software now?

AI is real now. It's available to anyone building software. What should change?

The common answer: add a chatbot. Put a copilot in the sidebar. The underlying product stays the same.

We think it should go further.

Every productivity tool starts with the same assumption: you do the work. The tool stores it.

Your calendar knows where you need to be, but not why. Your CRM has the contact but not the context. You're the one walking between them, holding the full picture in your head.

That made sense when software couldn't do better.

And when you can't keep up (the tasks pile up, the CRM goes stale, the weekly review doesn't happen) the tools don't just stop working. They make you feel guilty about it.

The guilt tax

7 tasks overdue
Some since last Tuesday
Weekly review not completed
3 weeks in a row
CRM contacts need updating
14 records stale
Inbox zero failed
127 unread messages

A different premise

The system should do this work. Guilt should never be a design pattern.

What we refuse to build.

Before describing what DailyOS is, it's worth being clear about what it will never become. These are design convictions.

× Another inbox to check.
× A dashboard you have to maintain.
× A system that needs feeding to work.
× A tool that counts your failures.
× A platform that holds your thinking hostage.

Three ideas

DailyOS is built on three convictions about how AI should change the relationship between people and their tools.

01

AI should produce.
People should benefit.

The traditional model: you enter data, the tool organizes it. The AI-augmented model: you enter data, and an assistant helps. Both still start with you doing work.

The AI-native model is different. The system reads your calendar, follows your email, cross-references your contacts, and builds intelligence about every meeting and relationship — continuously. You didn't ask it to. You didn't prompt it. You open the app and it already knows.

This goes beyond automation. Automation follows rules you write. Intelligence means the system understands what you need before you articulate it.

02

Guilt is a
design failure.

Streaks. Overdue counts. Red badges that grow. Every productivity tool uses some form of shame as a retention mechanism. The guilt creates obligation. But it also creates a relationship with your tools that feels like being managed by a disappointed boss.

DailyOS has no guilt mechanics. No streaks. No badge counts. No notifications about what you haven't done. It works whether you use it every day or disappear for a month. When you come back, your day is prepared. The system doesn't care about your engagement. It cares about your morning.

03

Your mind is
not a product.

Who owns the intelligence that a system builds about you?

Your CRM vendor knows your relationships. Your calendar provider knows your schedule. Your AI assistant knows your patterns of thinking. All of that knowledge, built from your work and your decisions, lives on someone else's server.

DailyOS stores everything locally. SQLite database. Plain Markdown files. On your machine, in folders you control, in formats any tool can read. We believe the intelligence a system builds about your life belongs to you. No business model should be built on holding it hostage.

~/workspace/
.dailyos/
briefings/
transcripts/
intelligence/
config.json

If you stop using DailyOS, you take everything with you. Every briefing, every transcript, every piece of intelligence the system ever built. It was always yours.

These ideas aren't specific to DailyOS. If AI can read and prepare, tools should stop asking humans to do that work. If value drives engagement, shame has no place in design. If intelligence can be stored locally, your thinking shouldn't live on someone else's server.

The technology is here. We're trying to use it well.

The system works.
You walk in ready.

DailyOS is an open-source project, currently in alpha for Automatticians on macOS.

Get started →
* * *
Built on three ideas. Tested every morning.