Vernor Vinge writes about the future, and the worlds he builds fascinate me. They’re packed with weird but plausible tools, and he understands the world-shaking power of communication. If you want a flavor of his writing, check out his first novella True Names online.

He’s best known for predicting the singularity, but I think his most interesting innovations are much more personal, and I see a lot of Defrag’s community starting to turn them into reality. Here’s how some of is ideas are moving towards the real world.

SM-ing

In Rainbow’s End, a lot of the dialog takes place by ‘Secret Messaging’, or sming. It’s like instant messaging, but with an interface through wearable computers that lets anyone silently send with no sign to the outside world. One of the most revolutionary uses of IM I found at Apple was having hidden side-channels to your colleagues in meetings, so you could frankly compare notes while someone else was talking, and plan your next moves together. This required a laptop open and visible typing though, but Vinge understands how much a truly hidden version of this changes. The plot is driven by people conspiring through sming, everyone’s addicted to it.

You can see how we might end up there with companies like Parlano turning IM into a vital workplace tool, and SMS through mobile devices making it possible to send from anywhere. The real leap will be the interface, and though it seems sending through something like tapping or subtle gestures should be possible, receiving requires retinal displays.

Relevance mining

Even way back in 1979’s True Names, Vinge’s background in computer science let him see how you could find patterns in massive sets of data. One of his recurring themes is that the characters have technology that searches through large sets of messages and pulls out a small group that are potentially interesting. This sort of assistance is a lot more plausible than the common SF trap of tools that require computers to grasp meaning, since that’s AI-complete. Having this kind of filter on your communications lets you monitor far more channels than you could without that help.

Companies like AideRSS are doing this already with their PostRank algorithm to pick out relevant articles from the noise of the blogosphere. You could even view Google’s PageRank as doing something similar for the wider web, and Xobni attempts to do the same thing for email. The key is that these are largely statistical methods that don’t need to understand the contents of a document at all.

Implicit Data

A lot of Vinge’s writing involves some kind of intelligence agency as a protagonist, and the use of advanced traffic analysis to reveal hidden information. Both traffic analysis and Google’s PageRank are part of a class of algorithms that use incidental, ‘implicit’ data to infer information that can’t be accessed directly.

Me.dium is trying to do something practical with this, building tools like search on top of massive amounts of data gathered on their user’s browsing habits, though you can argue that companies like Amazon have been doing it for years with their proprietary recommendation systems. The biggest barrier to wider adoption is how hard it is to access information across services or companies. To realize the promise of implicit data, we’ll need a lot more openness across the different silos that currently exist.