We need to shift the focus from 'AI risk,' whether 'existential' or not, to AI harm.
'AI' (automated decision systems, LLMs, generative tools for images, voice, and video, predictive policing and other forms of automated carcerality, and more) are harming people RIGHT NOW.