At one point, I caught myself planning the next twenty years instead of the next contract. I knew how to start another project. I was less sure which ideas deserved a large part of the working life ahead of me.
I have written software since 1993 and worked in commercial IT since 1997. Development led to architecture, team leadership, and eventually responsibility for the whole journey from an early idea to a working system. Plenty of tools made one part of that journey faster.
AI agents let me test a decision from two sides. One gathered the facts while the other challenged the direction I was considering. I sat down with both answers and made the call. Over time I began calling this AI-First.
I was still responsible for the choice. Now I could carry one idea much farther before bringing in more people. That turned a technical change into a personal question: which ideas deserved the next twenty years?
What to spend the next twenty years building
I wanted the technology to serve something I would still care about years later. Two ambitious ideas made the list, and I decided to make room for them now.
At the same time, software architecture and development still pay my bills and support my family. Those ideas would have to grow beside commercial work, which meant I would leave them for a while and come back.
Coming back was the expensive part. I might remember the choice but lose the reason behind it. I started building Arcanada so I could reopen a project, find the last decision, and pick up the thread with AI agents beside me.
What I wanted from Arcanada was ordinary: a record that told me where I had stopped and why. That need gave both long-term ideas a place beside the work that supports my family.