Remember poking on Facebook? It got a silly/meaningless semantic attached to it, likely because it was created before people could do it from their phone, and so it never went anywhere. You just poked people and then got poked back, ad infinitum.

Yo has exactly the same data model as Facebook pokes (a graph where the vertices are people and the edges are binary signal flags), but is different-looking enough that people are willing to grant it a different semantic interpretation. The semantic people seem to have adopted for a Yo is “hey, do the thing you know I want you to do, now’s the time” — which is, admittedly, a pretty useful semantic that was surprisingly missing from online interaction.

This semantic didn’t have to missing—someone could communicate the same semantic inherent in a Yo over Twitter, or SMS, or really any push-enabled app. But people will only understand services through the lens of one semantic at a time: you can't get someone to use Facebook as Twitter, or Wordpress as Tumblr, or Youtube as Coursera. Because nothing was just for this--nothing had this "rut" of suggested usage burned into its UX--users didn't bother doing this interaction at all. And now they do.

There's probably a market for quite a few other services that have effectively the same data models as apps we already have (Facebook, Twitter, Foursquare, Snapchat, Reddit, etc.) but which use different UX to push users into a new semantic interpretation of those data models. (To get different information from the same data.) HN seems completely color-blind to this idea, interestingly.