Summary:
- This article discusses the concept of "human-centered AI" and argues that it is a flawed and misleading idea. The author, Audrey Watters, a respected technology writer, suggests that AI systems are inherently designed to serve the interests of the corporations and institutions that create them, rather than the needs of individual humans.
- The article highlights how AI systems are often trained on data that reflects the biases and perspectives of the people who create them, leading to the perpetuation of societal inequalities and discrimination. It also explores how the rhetoric of "human-centered AI" can be used to obscure the true power dynamics and motivations behind the development of these technologies.
- The author concludes that rather than pursuing the illusion of "human-centered AI," we should instead focus on building technology that truly serves the needs and interests of all people, not just the powerful entities that control the development of these systems.