Predicting future is not that hard.

In 2001, guided by the invention principle I call "Telegrams before the train", my son and I came up with an idea of a portable playlist (US patent application 20040057348). This year, a new music service Songvote pushed the concept a little bit further:

...about 70 percent of music enthusiasts don't want to spend hours creating the perfect playlists, meaning that unless you're a DJ--or just really hard-core about music arrangement--you'd probably prefer to just listen to your library on shuffle or have someone else do the mixing for you.

Songvote is the service for you. This extremely nifty (and completely free) Web-based app lets users create collaborative playlists based on a simple time-restrained voting system. Using it is ridiculously straightforward: visit the site; create a theme-, event-, or activity-based contest; and then sit back and wait for the votes to roll in.

This "Telegram" principle is hard at work within any large-scale system that aims to optimize its performance. According to it, in order for the system to become more efficient, it has to shift upstream (to the Source and Distribution) the task of solving detection problems, e.g. determining the "quality" of a particular song. That's how it was done with the telegrams and train schedules 150 years ago; and this is how it is being done today with playlists and songs.

tags: detection, control, telegraph, example, course, payload