4:19 PM
@AaronMazel-Gee yeah, that's what I was thinking initially, but it was pretty hard to square with the fact that being tensored over finite spectra (or being stable) is a property of an $\infty$-category, just like being an additive (or abelian) category is a property of a category -- the enrichment is unique if it exists, so apparently it can't be twisted.
@EricPeterson I think this makes sense to me when I dualize everything. If I have a map $X \leftarrow Y$, then I get a simplicial object $(Y, Y \times_X Y, Y \times_X Y \times_X Y, \dots)$ -- for example if $X \leftarrow Y$ is an open cover, then this is the Cech complex. Then the isomorphism $(Y \times_X Y) \times_Y (Y \times_X Y) \leftarrow Y \times_X Y \times_X Y$ is the Segal condition corresponding to $\Lambda^1[2] \to \Delta[2]$.
Note that the inner simplicial structure maps are $S$-module maps, even though the outer ones aren't.
And tensoring over $S$ corresponds to the fact that the Segal maps being pullbacks over higher parts of the simplicial object than just the 0th one.
This is okay, because they're pullbacks over inner face maps.