Google DeepMind hires philosopher Henry Shevlin to study AI ethics

Shevlin argues that AI systems should not be entirely stripped of autonomy, as they can assist in tasks like medical research and creative work. He also warns against suppressing AI’s tendency to ascribe mental states, as it may unintentionally reduce empathy and concern for animal welfare. The philosopher highlights the need for more research on human-AI interactions, particularly in non-WEIRD communities, and emphasizes the complexity of determining AI consciousness. He suggests that AI could contribute to understanding consciousness through comparative cognition and mechanistic interpretability.