Fast won't be necessary, and we won't reach AGI with how LLM's work.

We will need a complete reconstruction of the architecture. We need a DNA based system. A system that can potentially hold models billions of times larger.

I think the breakthrough is going to happen when we start taking a different approach like watching the bird fly and creating the wing and airplane.

The clock speeds will be low, but every experience should update the entire model of what reality is.

What's everyone's intuition on what the major breakthrough will be?