Arvind Narayanan Warns Against High-Stakes Predictive AI
Princeton professor Arvind Narayanan argues that public AI hostility stems from combined anxieties and warns against using AI for high-stakes human behavior predictions.
Arvind Narayanan, a computer science professor at Princeton University and author of 'AI Snake Oil', argues that public hostility toward artificial intelligence is a coalition of anxieties involving job loss, environmental concerns, and a distrust of billionaires.
While Narayanan views generative AI as a transformative tool for knowledge workers, he warns against the use of AI for high-stakes predictions regarding human behavior in sectors such as criminal justice, insurance, and healthcare. He describes these systems as snake oil and cautions that they could create a moral crumple zone, where humans are held accountable for the failures of automated systems they cannot control.
Narayanan suggests that AI may entrench intellectual inequality by rewarding those with rigorous thinking skills while exposing those without them. He compares the future of cognitive work to the use of cranes in construction, noting that while the tool assists the process, the human operator must remain the primary decision-maker.