Classification, Slippage, Failure and Discovery

TL;DR

This paper explores how machine learning classification systems can serve as tools for constructive critique and discovery, highlighting their potential for revealing insights through their failures in image data and neural networks.

Contribution

It introduces experiments demonstrating the use of classification systems for critique and discovery, emphasizing the role of failure analysis in machine learning.

Findings

Classification systems can be used for constructive critique.

Failures in neural network classification can lead to new insights.

Experiments show potential for discovery through analysis of classification slippage.

Abstract

This text argues for the potential of machine learning infused classification systems as vectors for a technically-engaged and constructive technology critique. The text describes this potential with several experiments in image data creation and neural network based classification. The text considers varying aspects of slippage in classification and considers the potential for discovery - as opposed to disaster - stemming from machine learning systems when they fail to perform as anticipated.