Where the idea comes from
In Isaac Asimov's Foundation, a group of people works to preserve human knowledge and shorten the dark age that would follow the fall of an empire. What stayed with me from the saga is the idea of knowledge that survives because the people who come after keep using and developing it.
Technology got very good at creating new things and is still bad at preserving what we have already learned. Knowledge disappears when a person leaves, when a community breaks up, or when a platform shuts down. Recipes, techniques, decisions, and practices get lost in feeds and chats, without origin or context.
The thesis
I believe it's possible to build products where taking part leaves something for whoever comes next. A contribution stays linked to the person who made it, can be adapted by others, and reaches the next person with the history of what changed. The product gets better because the community learned, not only because it gained users.
I call this collective continuity: a community's ability to preserve, understand, develop, and pass on what it has built.
The role of AI
AI is an important tool for this. It helps find, connect, summarize, and translate what a community has produced. The condition is that it never erases the origin: an answer generated by a model does not replace the people who produced that knowledge.
How I decide what to build
Before becoming a product, every idea goes through four questions. If any answer is no, the idea doesn't belong at Longward, even if it's a good one. It can stay at fael.tech as a study or a tool.
- Is there a real community? Without the relationships between people, an essential part of the value disappears.
- Is there something worth preserving? The value goes beyond what the software does.
- Does the business grow with the community? Revenue doesn't depend on exploiting the people who take part.
- Does the origin stay visible? Authorship, context, and versions travel with every contribution.
How I'll build it
I'll build Longward's products with semi-autonomous agentic processes, using Radiant and other methods.
They may start free and add paid, premium, or enterprise plans. What I don't want is a business that needs to hold on to people's attention or their data to turn a profit.