The Autophonics method for music generation uses techniques from the field of Artificial Intelligence to closely emulate human decision making, which gives the resulting music a definite human feel.
Because of its algorithmic strategy, software applications built on the Autophonics method can be very small, taking up little disk space, thus making them suitable for use where size is a restricting factor, e. g. in mobile phones. They are also comparatively undemanding of processor power.

Traditionally, algorithmic generation of music yields esoteric, Avant Garde style compositions, of intrerest only to a small fraction of the listening community. Computer programs generating music in a more popular vein, normally uses either skilled listerner input or are data base driven, combining different sources of humanly made music into new compositions.
The Autophonics method changes all this, and makes it possible to build music generating software that work in any style - without the encumberment of a large data base and without being dependent on a musically competent person giving it input.
The music generation is easily guided by any listener, through rating each new song, which makes the upcoming song adapt to the listeners taste. The listener thus plays an important role in the creation of the music, which becomes more and more personalized. This empowerment of the ordinary music listener, to perform creative acts until now only available to composers and musicians, is one of the great strengths of the Autophonics music generation technique.

The core Autophonics algorithms are patented in Sweden, USA and China. More patents are pending.