I disagree, it doesn’t look confused to me.
Specifically on this (although possibly also on other stuff): (a) causal and statistical DAGs are fundamentally not the same kind of object,
The post explicitly discusses the different views of causality.
(b) no practical decision theory used by anyone includes the agent inside the DAG in the way this post describes.
That seems in line with what the post describes: “and so it will be (much) too large for me to reason about explicitly”.
I disagree, it doesn’t look confused to me.
The post explicitly discusses the different views of causality.
That seems in line with what the post describes: “and so it will be (much) too large for me to reason about explicitly”.