Laukkonen calls this a “metacognitive gradient,” a felt slope that guides action when we can't possibly compute the optimal path ourselves. Which we can't, because the effective complexity of reality vastly exceeds our modeling capacity.

If you can't model the whole, you can't derive the right action from first principles. You're forced to rely on... See more

Everything in nature is embedded. Cells are embedded in organs, organs in organisms, organisms in ecosystems. At every level, the part depends on the whole for its existence and function, and receives signals from that whole about how to behave. A neuron doesn't know where to fire without the context of the brain around it. A cell doesn't know... See more