The Fallacy of AI Functionality

What it is

Raji and colleagues argue many deployed AI systems fail on their own stated terms — they simply do not work — and that this belongs in the harm conversation alongside bias.

Topics