90% of us arguing about which model is best would not notice if you swapped them behind our backs

mild heresy for a sub that liveblogs every release. we spend enormous energy on which model wins which benchmark, Opus vs the new Sonnet vs whatever the other labs shipped this week, as if our daily work lives or dies on a few points of difference. and for a small number of people doing genuinely frontier stuff, it does. but be honest about your actual usage. drafting an email. cleaning up notes. asking a question a competent person could answer. summarizing a doc. if someone quietly swapped the model power

添加评论
点赞收藏
点踩分享查看原文
评论
?
参与讨论