Systems with very simple rules can produce unexpectedly complex behavior when scaled.
Ants — individually dumb, colony = intelligent
Neurons — individually simple, brain = consciousness
Transformers — simple next-token prediction = intelligence-like behavior
This is the idea of emergence.
Ada Lovelace saw that a simple computational substrate could represent logic and all forms of memory. Turing and his peers saw that logic itself could construct an initial state of mind — and a reward-driven process that enables learning.
The next step function in general intelligence will not come from a fictional design yet to be found.
It will come from a better arrangement of the building blocks already at hand.
We are engineering a core commodity that scales beautifully on its own — where and how it fits in the world are fundamental factors of its design.