Long before machine learning, nudges and dark patterns began fighting for our attention, the web was a welcoming and exciting place, where you could feed your curiosity by choice rather than by algorithmic routine.
Worse, since the arrival of LLMs, reading and staying informed seem to have become chores to get through as fast as possible. AI summaries, automatic classification and semantic extraction promise the essentials, but a summary is not what the author set out to say. These are symptoms of a wider problem, one that affects our ability to concentrate and to truly grasp a subject.
With Papercove, I’m trying to offer the opposite: a reader with a minimal interface, built on proven technologies, where you can read calmly. Here are the rules I follow every day:
-
Build with care.
Building software has never been this easy, and that tends to pull us away from our own products.
AI can help day to day, but what matters is still caring about the product, listening to the people who use it and experimenting with ways to solve a problem. So if building is no longer the hard part, caring should be.
-
Charge a fair price, and keep it steady.
Like many people, I've often watched apps raise their prices again and again while pushing users toward higher plans to improve margins, usually at the expense of the experience.
Papercove costs €48 a year, with a single plan. That price was designed to last, based on what maintenance and new features cost and on my ambitions for the project.
-
No ads, no freemium.
Papercove will never show ads or build a business model on its users' data. There's no tracking in the app to analyze behavior either: for that, I'd rather talk directly with the people who use it.
No free tier either. Usually fewer than 5% of a free app's users end up paying for it in the long run, which shifts the cost onto customers and makes it harder to offer good support.
-
Keep AI out of the product.
Papercove has no built-in AI assistant and no automatic article summaries, for a few reasons. First, our attention is slowly fading, and we shouldn't try to optimize everything we do all the time: reading should remain an end in itself.
Then, these features often mean compromising on privacy and security, or adding pricier plans based on the cost of tokens.
The only exception will be an MCP server (basically an API) for people who still want AI in their workflow, kept completely separate from the rest of the app.
-
Keep things simple.
I originally built Papercove because I love minimalist apps that can still do a lot, intelligently. I didn't always feel at home with current reading apps, which sometimes increase cognitive load even though a reader should be a peaceful place.
That's also why I will sometimes say no to features, which I think keeps Papercove stable.
-
Tread lightly.
Software can look like it leaves a light footprint, but its physical cost is real.
So I try to take a minimalist approach, both in choosing features and in building Papercove, to keep its environmental footprint as small as possible.
That’s it. It isn’t always perfect, but I do my best to keep improving on all of these points. If this is the kind of software you want to use, welcome aboard.
Robin