Language evolves, sometimes taking our most cherished nuggets of wisdom along for the ride. More than 250 years ago, the philosopher Voltaire wrote “le mieux est l’ennemi du bien,” the best is the enemy of the good. In the 20th century, this observation was further refined as “Perfection is the enemy of progress.” More recently, the popular dictum has been adapted in tech circles, to: “Move fast and break things.”
Nowhere is this core tenet more evident than in artificial intelligence. Large language models (LLMs) were trained on data scraped from the internet and rushed to market, pushing progress – and profit – ahead of perfection. The industry seems to be counting on updates to deal with pesky blips such as six-fingered hands and responses of pure fabrication framed as fact, generously