IBM Watson


For IBM Watson’s collaboration with Alex Da Kid, we created a real-time visualization software and installation. By interacting with a bespoke control console, the user of this installation could navigate through Watson’s vast analysis of song lyrics, news content and social media posts. Data queries could be filtered by time, emotional tenor.


My focus was on the implementing generative graphics. A wide number of parameters were used to drive the “color field” mesh (the basis of our visualization). I relied heavily on domain warping noise for the color bands representing the emotional tenor of the data. See Inigo’s article on the topic:

Twist, displacement height, noise speed, and post bloom and blur effects (using the fast Kawase method) were also driven by the Watson data.

ue4 previz

As a whole, this installation was created using include Cinder, Python, Arduino and two computers! Huge thanks to Hai Nguyen for the release of his signed distance field text Cinder block.

Role: Creative Developer