A digital DJ performing live

How Artificial Intelligence can maximize the energy of your party!

This post describes a patent application on how technology can optimize music experiences via understanding audience reactions and adapting — all powered by gamification techniques.

An intelligent, digital DJ performing live, through continuous interaction with its audience

Context: A crowded party; the AI DJ serves songs while capturing audiences reactions through (a) motion detection (mobile device based; also any wearable smart device like a watch or similar) (b) single or massive gestures (c) sound-based reactions (d) facial recognition-based patterns (e) concurrency-based patterns.

Starting by ‘trial and error’, the system starts serving songs to the audience; a good reaction to a song gives extra time to it and also guides the AI DJ on finding similar or related or associated songs (progressively increasing or maintaining audience engagement). A poor reaction makes the AI DJ to quickly replace it with something better.

The AI DJ optimizes song selection based on an audience response and according to a gamification process. The AI DJ receives data about audience members and determines a state and dynamic of the audience in response to particular songs played.

The AI DJ is able to identify audience leaders or laggards from gamification data or patterns about audience members. The gamification scores may be computed from the reactions or behaviors of audience members. The AI D J automatically adapts song selection logic based on the state and dynamic of the audience in general and/or based on the reactions of people with certain gamification scores.

Data relating states, dynamics, gamification scores, and tracks or sequences of tracks from previous events served by the AI DJ may help plan and optimize the songs selected and may be stored for planning future presentations. The AI DJ may adapt to audience preferences in real-time as controlled, at least in part, by gamification logic and related feedback loops.

The AI DJ may receive real-time feedback from sensors (e.g., cameras, microphones, accelerometers) positioned at a venue or carried by attendees and make determinations about a track or mix or the event itself, based on the feedback. The sensors may be hardware sensors or software sensors.

The AI DJ may base track and/or mix decisions on a combination of feedback from the audience as a whole and key members of the audience. A AI DJ may consider gamification patterns that enables identifying significant audience members (e.g., most socially relevant attendee) and weighting their reactions to a track more heavily than less socially relevant attendees.

The AI DJ may have a party timeline that establishes a path and trajectory for the mood at different points in a party. An AI DJ may therefore select tracks in sequences for the mix that will produce peaks and valleys in dance energy and dancer volume (e.g., number of people dancing). Additionally, the party timeline may be crafted for different demographics at different times. For example, a mix may be crafted so that a first demographic will dance followed by a second demographic.

While one demographic is dancing the other demographic may have opportunities for other interactions and vice versa. Since the AI DJ may want to follow a party timeline, tracks may be selected based on their interactions with other tracks.


Example apparatus and methods may employ facial recognition to further enhance the experience produced by an AI DJ. An AI DJ may identify as many faces as possible in the audience and may then try to produce a mix that gets a threshold number of identified faces to react in a certain positive manner.

Facial recognition may also be used for a finer grained evaluation of a track. In one embodiment, the facial analysis may produce metadata about audience members.


One clap, two clap, three clap, forty?

By clapping more or less, you can signal to us which stories really stand out.