The Myth of Ethical AI

The provided text, an excerpt from the article "The Myth of Ethical AI" by philosopher Bry Willis, argues that the concept of "Ethical AI" is fundamentally flawed and oxymoronic, akin to phrases like "compassionate capitalism." Willis contends that ethics is not a unified or quantifiable feature that can simply be programmed into machine learning systems, citing the existence of various, often contradictory, ethical frameworks. The author critiques the prevailing Silicon Valley approach, which relies on a "bland utilitarian consequentialism" that masks inherent biases and attempts to treat moral worth as a dataset. Furthermore, Willis suggests that the industry's focus on "alignment" is merely a euphemism for obedience to the values of the tech elite and a means to preempt regulation. Ultimately, the article posits that AI systems merely reflect and optimise the existing, often questionable, ethical values—such as profit and efficiency—of their human creators, concluding that the focus should be on creating ethical human architects rather than demanding impossible morality from machines.https://philosophics.blog/