we're stuck in this perpetual cycle of AI eniasm where every 5 years we re-discover the idea of machine learning, remake the same old graphs and papers from 10 years ago, and claim breakthroughs where there are none.