The AI could really just: - select the best next song given a simple input (for example, a microphone or a camera looking into the crowd) - mix it into the cu

The AI could really just: - select the best next song given a simple input (for example, a microphone or a camera looking into the crowd) - mix it into the current song - repeat