QUANTUM FRAME
LITTLE SOUND MACHINES
LITTLE SOUND MACHINES
2019 -
2018
2018
The Quantum Frame is a mechanical installation that speculates on the future of quantum computing and what that may hold for machine intelligence and consciousness. The installation takes the form of the present day quantum computer, with a tubular central chamber, where machine learning data drives the mechanical movements of an electromagnetic structure, breathing life into the metallic framework, a ghost in the machine.
The current version of the frame is self-generative. But the artist hopes that once time-sharing of the quantum computer is open to the general public, that the installation may be able to talk with the quantum machine directly via data transfer.
This piece is currently on-going. Magnetic field experimentation and research with ferrofluid are currently in--progress.
The Little Sound Machines is a sound installation consisting of a series of both mechanical and digital machines that are connected to an AI network. Three AIs form the central brain of the network. Through learning from and influencing each other, the AIs construct the musical phrases that are then played out through a series of sound-generating machines. The music generated by the AI is also presented on a series of television screens that visualizes both the AI data and audio, as well as machine logic and behavior to the audience.
This piece proposes a new mode of music creation in the age of intelligent machine. Through experimentation, the artist presents an exploration of new musical interfaces that erases the composer from the equation, to present a purely machine-made performance.
The Little Sound Machines are made from found objects, up-cycled and spare parts.
The Little Sound Machines is a sound installation consisting of a series of both mechanical and digital machines that are connected to an AI network. Three AIs form the central brain of the network. Through learning from and influencing each other, the AIs construct the musical phrases that are then played out through a series of sound-generating machines. The music generated by the AI is also presented on a series of television screens that visualizes both the AI data and audio, as well as machine logic and behavior to the audience.
This piece proposes a new mode of music creation in the age of intelligent machine. Through experimentation, the artist presents an exploration of new musical interfaces that erases the composer from the equation, to present a purely machine-made performance.
The Little Sound Machines are made from found objects, up-cycled and spare parts.
LITTLE SOUND MACHINES
2018
The Little Sound Machines is a sound installation consisting of a series of both mechanical and digital machines that are connected to an AI network. Three AIs form the central brain of the network. Through learning from and influencing each other, the AIs construct the musical phrases that are then played out through a series of sound-generating machines. The music generated by the AI is also presented on a series of television screens that visualizes both the AI data and audio, as well as machine logic and behavior to the audience.
This piece proposes a new mode of music creation in the age of intelligent machine. Through experimentation, the artist presents an exploration of new musical interfaces that erases the composer from the equation, to present a purely machine-made performance.
The Little Sound Machines are made from found objects, up-cycled and spare parts.
NEURAL DREAM
VIDEO SERIES
ML, video
2022
Neural Dreams / Human Machines is a series of works that the artist began in 2014 that experiments with the Google Deep Dream neural network as a tool and approach for generating computationally processed imagery. Utilizing public domain video footage, television advertisements and scraped video footage from the internet, this series explores how machines translate and interpret human mass media and visual culture. This process utilizes a method of frame-by-frame image processing with assistance of open-source ML tools. Found visual content is processed through multiple generations and layers of inception and pieced together manually with an emphasis on exhibiting the image morphology process. This work experiments with early day machine learning tools and neural networks in a figurative and material way, where custom programming was necessary and long processing durations spanned days and weeks.
82 Layers of Inception, 2022 (33 minutes, 40 seconds looped)
Original Kinder Dream Inception, 2015; 82 Layers of Inception and Multiple Generations, 2022
82 Layers of Inception is a reinterpretation of the original Kinder Dream experimentation with Google Deep Dream in 2015. Reusing the original Kinder Surprise Chocolate Egg Commercial: Humpty Dumpty TV commercial from 1982. The footage was passed through Google Deep Dream 82 times, with each pass being processed at a different layer of inception. The 82 videos are presented in a grid of 9x9, each representing one layer of inception. When viewing, each sequential layer is masked over the grid, rotating through the entire sequence of videos.
Original Source Advertisement: Kinder Surprise Chocolate Egg Commercial: Humpty Dumpty – Lansdowneuro, 1983-1984.
This work is currently exhibiting at Museo Del Hongo: Holy Childern in Berlin, Germany, September 2022.
Ice Cream Dream, 2022 (10 minutes, 48 seconds looped)
Original Ice Cream Dream Inception, 2014; Re-Imagined through 5 Layers of Inception and Multiple Generations, 2022
Ice Cream Dream is one of the first experiments processing found TV commercials from the internet through Google Deep Dream neural network in 2014. It was processed 4 times through a single inception layer. In 2022, the original 2014 Ice Cream Dream video was used as source material and processed through 5 layers of inception and a computational drawing algorithm.
Original Source Advertisement: Little Baby’s Ice Cream “This is a Special Time” - Doug Garth Williams, 2012.
82 Layers of Inception, frames from video, 2022 (33min 40sec)
Ice Cream Dream, frames from video, 2022 (10min 48sec)