AI Ethics Reading

Do you find yourself preparing for the new world of AI? Are you learning how to prompt? Are you trying to figure out how your job will change and what skills you will need to have? You should be.

But you should also understand the ethics of AI. Where is this data coming from? Who decides what labels are used? What decisions can tolerate the mistakes of AI? What resources are we using to build and power the data centers running AI workloads? Who benefits and who suffers?

Here are some books to check out on the topic:

Army of None: Autonomous Weapons and the Future of War by Paul Scharre
AI fears can sometimes feel like an existential threat. It is much more real when autonomous weapons are designed to kill. If we no longer risk human life in our war efforts, will we engage in war more freely?
Atlas of AI: Power, Politics, and the Planetary Costs of Artificial Intelligence by Kate Crawford – An inconvenient truth for the technology industry in general and AI in particular. Who is paid to label data? Where do the raw materials come from to build the hardware to run AI? Who holds the power to make these decisions?
Unmasking AI: My Mission to Protect What Is Human in a World of Machines by Joy Buolamwini
The coded gaze. Already oppressed groups can be further excoded by AI systems. Neutral does not mean unbiased. Why should we trust closed AI systems let loose on society?
Life is a Miracle: An Essay Against Modern Superstition by Wendall Berry
This is not an AI book and was published in 2001, but it is relevant today. Why should we always assume technical advances cause progress and not regression? Should we do everything science is able to do? Who gets to decide?
Driverless: Intelligent Cars and the Road Ahead by Hod Lipson and Melba Kurman
If an old lady steps in front of an autonomous vehicle, does it swerve into a crowd of young pedestrians to avoid her? A human driver could be forgiven for either choice made in the moment. It feels much different when we have time to encode our ethical choices in advance.