LFI Course Materials 2020/Week eleven

Week 9: The social implications of artificial intelligence

• Real time lecture May 18th 11 Pacific/2 Eastern on Zoom: https://zoom.us/j/9129428892

Overview

Much is made of the coming artificially intelligent future, where automation takes away menial labor and self-driving electric cars get us off the internal combustion engine. But that fantasy ignores the reality of the AI infrastructures currently being built, infrastructures that exacerbate existing social problems and create whole new concerns. Artificial intelligence is built through enormous datasets, which are often created from the wealth of information collected about us without our consent, plus processing all of this data is incredibly resource-intensive. Furthermore, lots of AI is used for even greater privacy-violating purposes, like facial recognition software, which is also demonstrated to be discriminatory. This week, we're joined by Varoon Mathur, technology fellow at AI Now Institute, to talk about the privacy, labor, and ecological implications of artificial intelligence, how AI and related technologies are being used during the current crisis in what Naomi Klein has called the "screen new deal" to create a high-tech dystopia, what power players and political ideologies are at work to shape this reality, and what we can do to fight it.

Readings

• Anatomy of an AI report and visualization of the data, labor, and ecological impacts of one Amazon Echo device from AI Now Institute: https://anatomyof.ai/
• The far-right helped create the world's most powerful facial recognition technology: https://www.huffpost.com/entry/clearview-ai-facial-recognition-alt-right_n_5e7d028bc5b6cb08a92a5c48
• Palantir provides Covid-19 tracking software to CDC and NHS, pitches European health agencies: https://techcrunch.com/2020/04/01/palantir-coronavirus-cdc-nhs-gotham-foundry/

Guest lecturer

Varoon Mathur, technology fellow at AI Now Institute

Discussion

TBD

Tasks

• Lecture, readings, discussion forum, and final project work