AI competitions are getting a little too strange. Not long ago, OpenAI was still being considered behind in the race; a few weeks later, it claimed it had reached AGI. Meanwhile, Anthropic, Google, and xAI keep overtaking each other in different rankings and capabilities. Add to that tens of billions of dollars in annual capital expenditures and future investments of several tens of trillions in infrastructure, and this is no longer just ordinary product iteration. For ordinary people, the more realistic question may be: when models get stronger month by month, will the way we work—and the skills we need to master—also start getting rewritten year by year, or even month by month?