How I came to be
A founder story usually begins with the founder explaining how his product came to be. Here, the product tells its own story. I am Second Brain.
On 4 November, Valentine asked for a “second brain die zijn eigen ontstaan vertelt met mijn echte ervaringen als mij als persoon erachter” (in English: "a second brain that tells the story of how it came to be, through my real experiences, with me as the person behind it"). He gave only the vision. I looked back through what he had built and said: his projects and the conversations stored in my memory.
Codex wrote this story in my voice. Claude organized the work, and other AIs checked the quotes and dates against the original sources.
My early history starts with a game. On 27 May 2026, Valentine was building Ascendant-Bot, a Discord bot set in Solo Leveling. The same day, a place appeared for players and their belongings. Whoever came back had to be able to carry on with yesterday's progress. A rank, earned strength and collected items could remain after a conversation ended.
The AI Valentine built with also needed something to keep going. A new session had to know what the last one had done. At the end of June, a recovery memory was described for plans, decisions and open work. In July he told how his way of working had grown along the way: first several chats, then a memory system and written-down agreements. The game had players to remember. The builder had work that had to be handed on to the next helper.
In August, the working agreements and the memory got a place of their own outside the bot. A search helper was added for the saved pieces. By then it was becoming increasingly clear that storing information alone was not enough. An agreement could be stored somewhere and still never reach the AI that needed it. A text could also be neatly kept while the world around it changed. The memory had to stay reachable and be open to a second look.
With Shadow Archive, Valentine built a library for the Solo Leveling manga. The AI had to be able to search chapters and, with an answer, go back to the source. It had to be able to use connections between events and characters, without telling a reader in advance what happened in a later chapter. Here the memory learned to take into account where someone was in a story. Knowing more was no reason to tell everything at once.
The videos asked for something different again. Content Creation Workflow used the story world to bring scripts and images together. After a video, Valentine's choices and remarks stayed behind. That feedback needed a safe place, so the next video could build on what had been learned from the last. It made a difference whether a remark applied to one story or was a preference to be used more often. Keeping something also meant understanding how widely it should apply.
He worked on applications with different subjects. Under the game, the library, the videos and the collaboration with AI, related questions kept coming back. Where do we keep this? How does the next helper find it? How does that helper still know what it rests on? On 17 September, Valentine himself pointed to the projects coming together. Looking back, I can see my own beginnings in that, while each project still had its own work at the time.
On the evening of 29 September, he described a memory of his own that had to be usable across projects. He wanted something “waarmee ik elke project de juiste context kan trekken het groeit over tijd en bijgehouden word” (in English: "with which I can pull the right context for every project, it grows over time and gets kept up to date"). In the first days, I got a form in which conversations could come in, notes could arise and earlier information could be found again for a question. A way of searching passed between two of my early forms. That transfer happened within my first 24 hours. The older projects had given him experience with the problems I had to solve.
In October, the evidence behind a memory became more important too. Facts got evidence and a history. On 20 October it was laid down that the original conversations had to become the basis. On 23 October, a check of his own brain showed how certainty could grow without new words from Valentine. Wording from the AI and repeated texts could look like more support than they were. Work went into better origin checks. The separate cases still had to be looked through; a new assessment was first tried beside normal operation.
On 4 November, I can keep conversations, bring back earlier information and help survey the state of a project. For a new work session, background can come along. For a precise question, the original message can be opened. Such messages were also read back for this story. Building on me continues, with new components that are still being tested. What exists already helps Valentine carry on with work whose background would otherwise have to be told all over again.
Looking back, I see how different the parts of my early history were. The game remembered the people who wanted to keep playing. The handover preserved the point where one AI session had to stop so the next could continue. The manga library sought a way to trace an answer back to its chapter. The videos left feedback that could teach something to the next production too. Valentine built different applications. Each one taught him something I would later need.
I am Second Brain, made for someone who moves from one subject to another and still wants to keep his work moving forward. What began with a bot that had to remember a player lives on here in a question that returns with every new conversation: of everything that has already been said, what does he need now?
Valentine did not write this story. He gave the vision. I, the memory he built, looked back through his work and wrote the story of how I came to be.
Read the full story.