Hand Gesture

Teaching Practice

For a group show, Joshua Moreno invited artists to share work they make as a result of teaching for a show at 120710 Gallery in Berkeley, CA. His prompt reminded me of a project I did to demonstrate how to make data-driven animations.

Kill the Messenger

The project began when the President fired the Secretary of the Bureau of Labor Statistics because the data she reported did not match the Presidents expectations in June of 2025. The classic Locus of Control fallacy is that we seek to gain control of something uncontrollable by over-correcting something controllable in its place. I collected the jobs lost and gained data from the five greatest economic hubs in the United States from the BLS website, and used the month-to-month changes as input to animate each finger of a CGI hand. I used the monthly Gross Domestic Product of the US to move the hand on the vertical up axis. The result is a “hand gesture” that articulates the constant tremor of the labor market, and the personal tremor of the people who lose or start a new job. In June of 2025, the data as we know it ends and the fingers animate into a smooth gesture of ideological alignment.

Original Prompt: Driven By Data

Select a time series with at least 100 data rows. Store the data locally as a comma-separated-value (.csv) file. Select a blender model, either your own creation, or an existing one, that relates to the subject of your data, and use the data in its normalized form to animate your model. Make sure the animation is no longer than 600 frames, but no shorter than the row count of your data.

Here is a sample blender file of the spheres controlled by labor data. It has a Python script to keyframe the z-positions of five spheres with labor data. To make all this work, you will need to edit the Python script to match your data and your data storage location. 

Here is the data that drives the animation as a .csv (comma-separated value) file.
It's sourced from the Bureau of Labor Statistics: https://www.bls.gov/regions/economic-summaries.htm. This is indeed the data over which the director of the Bureau of Labor Statistics was fired, because the posted data did not meet the Presidents' expectations. 

Data usually comes in forms that are not clean enough for animation. The original data here for example has letters mixed in with the numbers, and we must have only numbers to animate. To clean the original data from the Bureau of Labor Statistics, we can use the Jupyter Notebook script linked here.