Markov-chain-based musical creative intelligent agent passing successfully the Turing Test

Nowadays, the multimedia industry needs a quick generation of large volumes of new music content. For this purpose, we present in this paper an efficient solution for automated music composition based on Artificial Intelligence (AI). In particular, the developed approach integrates multiple machine learning algorithms based on Markov chains (MCs) within a system comprising a user-friendly interface for intelligent human-computer interactions, leading to a full, creative intelligent agent which is able to produce Musical Instrument Digital Interface (MIDI) format files with new, original music based on monophonic input seed music provided by the user. This AI-based music composition agent goes beyond algorithmic music generation, since it autonomously generates new pieces of music with a balance between harmonious musical form and non-deterministic novelty. Indeed, after passing through all the testing levels, this creative software agent has been deployed in real-world conditions and passed successfully the Turing test.