2025-05 Hey, just write and things will be eventually okay

I’m still working heavily on the most difficult project I tackled so far: the project of a how to research. I see now, it’s incredibly difficult to maintain a system where critical thought, discussion, is incentivized. Just intaking information does not do any good; this is the realization I come to over and over again.

I want to contribute meaningfully to the world. Contribute to technologies that I’m convinced will change the world. I find the area of safe autonomous systems essential for this — and I can only do a good work here if I think out of the box. Emergent ideas.

So my systems should foment the creation of emergent ideas. I don’t want to make this piece a piece about how to achieve this, since the answer is still unclear to me. I do however think that I made some progress during these couple weeks.

I realized I put so much weight on details: 2025-05 On naming notes is the living prove. Of course, working on this project involve exploring the whole field of possibilities; nothing wrong here. However, experience has shown to me, that if you just work on a project — let it kneed itself — you will end up on systems that conformed best (or best enough) to the task at the hand.

Best in this case is heavily debatable. Normally what ends up happening is most friction points are filed out; if there’s an allowed path of least resistance it’ll get chosen eventually. And this is my grand master plan behind researching: I have to do it long enough. It’s okay, it’s fine if everything is slow, sometimes redundant in this first stage. Let it sit and do it long enough. You will find what works and what not.

Nevertheless, this can only work if the optimization goals are clear from the start. These are (1) ideate and discuss: see the gaps and generate novel solutions for standing research questions; and (2) extract artifacts: a repository build in such a way that you can just “pluck” articles, posts or any deliverable. It should benefit from compound effects — pay us interest.

I just want to get these goals right from the start. As long what I am doing are aligned with that and my gut feeling say that what I am doing is correct, then keep it. If I have a hunch that something can be improved or would lead to a goal being deviated: discuss it. Write about, remember Writing is Thinking.

I guess this is what I wanted to hear, since I got a mind-freeing feeling after writing this. The task at hand is daunting. It’s enormous and open-endend: I understand why it can be overwhelming. However, you have implemented a system that is already providing utility; and you have also a support net if things go south.