Ep. 50: Talking Algorithms and Algorithmic Systems

Algorithms and Anthropology: In this episode, the focus is on understanding what algorithms are and how they relate to emerging technology. It introduces the idea that algorithms are a part of our daily lives, especially in the financial services world.

Defining Algorithms: Dr. Nick Seaver, an anthropologist at Tufts University, explains that an algorithm can be viewed in different ways—it can range from a simple set of steps to a more intricate machine learning software process that learns from source data and makes future predictions. This discussion also touches on the concept of algorithmic systems.

Complexity and Human Involvement: The episode highlights the complexity of algorithmic systems and how they differ from traditional algorithms. Nick Seaver emphasizes that understanding algorithms requires studying the people who create and maintain these systems, as well as the multitude of decisions that humans make within them. Human involvement and decision-making play a significant role in shaping algorithmic systems.

Culture and Algorithms: While exploring the role of culture in algorithmic systems,anthropologists like Nick Seaver are interested in how technical-minded people deal with cultural problems. Algorithms are not immune to cultural influence, and engineers building these systems often incorporate their own perceptions of culture into their designs. This can lead to various approaches and assumptions about why people like what they like.

Challenges and Bias in Algorithms: There are challenges in creating unbiased algorithms. In particular, the podcast explores the issue of proxy problems, where algorithms can inadvertently reproduce social biases, such as racism, even if they are designed to be race-blind. The discussion emphasizes the importance of ongoing vigilance, learning, and questioning within the development of algorithmic systems, especially in industries like financial services.