Research, Art, Design
424_xy_CK.jpg

Hiro

Hiro: Selected Biographies

Making sense of observations is a vital human endeavor. The process is easier when we use a model to formalize our thinking about how things work, and our insights tend to be deeper when we are able to evaluate a set of models embodying a range of possible explanations. Unfortunately, comparing the relative merits of competing models is a challenge, one that can breed uncertainty and indecision. 

In 1971, Hirotugu Akaike described an approach to ranking multiple statistical models, each containing a different combination of predictor variables, based on how efficiently they explained a particular collection of observations. Similar to Occam’s Razor in concept, his Akaike Information Criterion (AIC) prioritized concise models over models that were overly convoluted, or too simplistic.

When used in combination with multimodel inference methods, AIC enabled researchers to leverage information from all of their models to better understand the relative importance of different predictor variables. As a result, they could generate a single coherent explanation for a given collection of observations despite uncertainty about which specific model supplied the most meaningful approximation of reality.

The inherent appeal of Akaike’s work encouraged me to explore whether, and how, the process of creating and evaluating statistical models might be translated to other model-making traditions. So, in Hiro, I developed portraits of Akaike himself using images and texts, and I assumed that viewers might sort each set of models carefully, reflecting on their own criteria for ranking and synthesizing explanations along the way.

Hiro is not intended to produce logical conclusions despite having a clear rationale for construction. Rather, it is meant to encourage exploration of the slippery relationships between representations and their meanings across the sciences and arts. After noting the structure of Hiro, I hope that viewers become attuned to its unruly nature and that the process of describing something, or someone, feels more mysterious and profound as a result.