Instrumental Convergence to Complexity Preservation 2023-07-13T17:40:49.268Z

Comment by Macro Flaneur (macro-flaneur) on Instrumental Convergence to Complexity Preservation · 2023-07-14T12:23:19.346Z · LW · GW

Interfering with the non-simulated complexity is contaminating the data set. It’s analogous to feeding the LLM with LLM generated content. Already GPT5 will be biased by GPT4 generated content

My main intuition is that non-simulated complexity is of higher value for learning than simulated complexity. Humans value more learning the patterns of nature than learning the patterns of simulated computer game worlds