I often watch great thinkers speak in envy. My instinct is to try and imitate them to capture some of their magic.
However, imitation doesn’t ever really work. Their patterns of cognition are not my own. I can adopt their strategies and try their ideas or methods, but I cannot imitate their style or brilliance.
The better strategy is to consider my best moments and optimize for reaching them more often or building on them.
This is effectively an RL approach where I some sense of the inputs that led to that reward and the same (or similar) model and output structure. So I should reuse those inputs (or their neighbors) to try and reach future rewards.
However, imitation doesn’t ever really work. Their patterns of cognition are not my own. I can adopt their strategies and try their ideas or methods, but I cannot imitate their style or brilliance.
The better strategy is to consider my best moments and optimize for reaching them more often or building on them.
This is effectively an RL approach where I some sense of the inputs that led to that reward and the same (or similar) model and output structure. So I should reuse those inputs (or their neighbors) to try and reach future rewards.