Every app you use takes a copy of you.
Your browser keeps your history. Your notes app keeps your notes. Google keeps what you searched. Spotify keeps what you listened to. Pocket keeps what you saved to read later, until Pocket is acquired, pivots, or shuts down. Each of these companies isn't storing "your data" in any meaningful sense of the word "your." They're building records of you that you happen to have access to, for now, under their terms.
You are a tenant in your own digital history. You furnished the apartment. You made it yours. But you don't hold the deed.
The Problem Is Structural, Not Personal
Most people feel this problem individually but don't recognize it as structural. You've experienced the symptoms: a bookmarking service you relied on shutting down, years of saved articles gone overnight. A notes app changing its pricing model, forcing a choice between paying more or exporting your data into a format that nothing else reads cleanly. A social platform's algorithm burying things you used to easily find. Browser history that Chrome keeps but can't search by content, only by URL.
These feel like isolated frustrations. They're not. They're all the same underlying problem: the architecture of the modern web was not designed for you to own anything.
Every app you use is built on the assumption that your data should live on their servers. This isn't malicious. It's the default shape of how web applications get built. You log in, the app runs, your data accumulates in their database, you come back later and retrieve it. That model is convenient. It's also why you never truly own what you create or collect.
The numbers make this concrete. A 2024 Pew Research study found that 38% of webpages from 2013 are now inaccessible. A separate analysis by Ahrefs found that 66.5% of links from the past decade are now dead. The web is dissolving faster than most people realize, and the private records you kept using web-based tools are dissolving with it.
We Already Solved This for Money
Here's what makes the architecture problem so striking: we solved it for money over a century ago.
Your bank account is yours. Venmo can read from it and write to it, with your permission. Your payroll processor deposits into it. Your credit card company draws from it. Your financial software syncs it. None of these services "own" your money. They're interfaces to something that belongs to you. When you stop using Venmo, your money doesn't disappear. When your credit card company gets acquired, your balance transfers intact. The money is yours because the architecture was designed to make it yours.
We don't have anything like this for your digital life.
There is no equivalent of a bank account for what you've read, what you've searched, what you've watched, what you've bought, what you've thought, what you've saved, and what you've decided. Instead, each app accumulates a fragment of you in a proprietary silo, and you have no claim on the whole. Your reading history is in your browser vendor's servers. Your notes are in Notion's database. Your bookmarks are in Raindrop, or Pinboard, or the ghost of Delicious. Your highlights are in Readwise. Your research is in Google Drive. Taken together, these fragments would tell a rich story of who you are and what you know. Separated across a dozen services, they're just records someone else holds.
What a Different Architecture Looks Like
The inversion is straightforward to describe, even if it's technically non-trivial to build.
Instead of bringing your data to apps, apps come to your data.
You have a personal store. It lives on your device, not on a company's servers. When you install a notes app, it requests read and write access to your store, the way an iOS app requests access to your photo library. The app is an interface. Your data is the thing. When you delete the app, your data doesn't go with it. When you switch to a different app, you don't start over. When the startup that made your favorite tool runs out of money, your records survive.
The simplest version of this starts with two things that are actually the same thing: your browsing history and your intentional saves.
Every bookmark, highlight, note, or saved article is just a history item you decided mattered. Both start with the same root: you encountered a piece of content, at a moment in time, and something happened as a result. The passive record and the active save are not two separate systems that happen to share URLs. They're two points on a spectrum of the same relationship between you and an idea.
Call this a Web Memory. You visit a page; a record is created automatically. You visit again; the same record accumulates the return. You bookmark it; the record gets a flag. You add a note; your thinking attaches to the source. Over time, the record becomes richer. When you search later, you find everything at once: the content, the context, your annotations, the pattern of when and how you engaged with it.
This is not a new category of app. It's a new layer of infrastructure, one that sits beneath the apps you already use.
What's Standing in the Way
The reason this hasn't existed at scale isn't technical. The protocols and tools to build it are increasingly mature. The obstacle is economic and habitual.
Convenience is real. Google Docs is easier than owning your documents. The friction of a self-owned store, however small, competes against services that are genuinely seamless. Every prior attempt at this architecture, and there have been several serious ones, underestimated how hard it is to make personal data ownership feel as effortless as personal data rental.
There is also the bootstrapping problem. A personal store with no data in it is worthless. The history and saves you've already accumulated are scattered across a dozen services. Until there's a clean way to import that history, the new system starts from zero while the old systems start from years of accumulation.
And then there's the question of who builds this and how they sustain it. Protocols that nobody controls are hard to fund. Infrastructure without a business model attracts enthusiasm and then runs out of runway. The history of this space is littered with projects that had the right architecture and the wrong economics.
Why the Conditions Are Changing
Three things are shifting that weren't true five years ago.
First, the category of local-first software, tools designed so that your device holds the primary copy of your data and the cloud serves only as a transport layer, has gone from an academic fringe idea in 2019 to a growing ecosystem with production-ready tools and real user demand.
Second, the AI moment has changed the stakes. An AI assistant that operates over your actual memory, what you've really read and thought and decided, would be qualitatively more useful than one limited to what you've chosen to share with a third-party service. The value of a personal store isn't just preservation. It's the foundation for a genuinely personal AI. That reframes the whole category.
Third, and most importantly: people are increasingly aware that they're the product. The appetite for personal data sovereignty has moved from privacy advocates to mainstream users. The question is no longer whether people want to own their data. It's whether the alternative is convenient enough.
The Direction
The shape of what comes next is visible, even if the specific form isn't settled.
A personal store for your digital activity, starting with the browsing history and intentional saves that make up your web memory, is the minimum useful unit. It's concrete enough to build, familiar enough that users immediately understand the value, and foundational enough that more powerful capabilities can grow from it.
The architecture that will work is not a new app. It's a new layer. One that existing apps can read from and write to. One that users own rather than rent. One designed to outlast any individual service that builds on top of it.
We built this for money because the alternative, banks that could simply decide your balance no longer existed, was obviously unacceptable. The case for building it for your digital life is the same. The alternative, a digital memory that disappears when services change their terms, is unacceptable. We've just gotten used to it.
That's the opening. The question now is who fills it.
The Sovereign Memory Series
Part 1: You Don't Own Your Memory
Part 2: Your Reading History, Reimagined
Part 3: The Graveyard of Good Ideas
Part 4: The AI That Actually Knows You