"Ads Don't Work That Way" still a great essay but this example really didn't hold up. @KevinSimler have you reevaluated any of your ideas here since?pic.twitter.com/KTwAICRa8t
-
words are the worst :(
but yeah, i really wish there were better patterns for how to talk about these kinds of domains. “superimposed distributions” kinda gets at it.
-
-
Yeah words make everything so much more extreme. It's really hard to talk in terms of stable patterns, distributions, normative behavior, etc
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
I often think in terms of trying to decompose a high-dimensional space into ~principal components, and then describing the principal components in isolation. E.g. in social situations, there’s the prestige component, the dominance component, the coalition politics component, etc.
-
Each is best understood by thinking about it by itself, as a pure vector unadulterated with the other components. But then any real-world example is a data point that mixes everything together. And of course the vectors aren’t orthogonal, and they mix chaotically....
- 8 more replies
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.