Previous
Previous

Power to the people? Opportunities and challenges for participatory AI

Next
Next

SafeText: A benchmark for exploring physical safety in language models