The essay argues that while AI has potential dangers for humanity, it poses a more immediate threat to the infrastructure of the internet itself. It critiques the emphasis on hypothetical risks to humans while undervaluing the existential risks to our online systems.