I dream of a world where we don't ignore our problems but instead find the root issue, seeking to understand the world as precisely as we can. That's what we are trying to do at @BuildCoherence

Natural language has increasingly approximated formal logic over time. The end state is that the base language is formal logic.

Step-by-step reasoning is great until you realize it's just confident hallucination with annotations.

Axioms in physics are empirical, approximate, and judged by their position on the compression/accuracy frontier. Axioms just let people reason and expand their knowledge without shifting definitions.

We need to create a Moltbook on Axiom ASAP! Pure formal logic machine-machine communication. Ooooooo

Whether AI has different values from us is not the question, but whether or not AI has positive-sum values. Therefore, the most important pursuit is discovering and proving the set of positive-sum values.

AI alignment is not a real problem. If alignment means having human conceptions of x, and humans hold opposing conceptions of x, it is an impossible pursuit. Building a strictly coherent AI with respect to an explicit value system is however possible.

Moreover, I don't think monumental-ness is of concern. From the perspective of unprecedented AI development, population, social, and economy challenges, "monumental" may be *necessary.*

Is it true that for all more powerful entities, humanity loses? I don't think so. People who are disproportionately very intelligent have the desire to colonize mars before they want to wipe out primitive tribes. There's no utility to oppress people when you can go into outerspace

Even if AI takes over we will rely on AI to solve this problem.