Superhuman 的 AI 草稿 60% 无需修改 —— 但我们仍无法完全信任它
电子邮件客户端 superhuman 推出了一项新的自动草稿功能，使用了来自 openai 和 anthropic 的前沿模型。该公司声称 60% 的发送草稿不需要编辑，尽管准确性仍然不稳定。
email client superhuman released a new version of its auto-draft feature, utilizing a mix of frontier models from openai and anthropic to generate replies that mimic user tone. beta testing data indicates that 40% of drafts were sent within 24 hours, and 60% of those were sent without any edits. the system learns from user behavior over time to correct mistakes, such as refusing unreasonable meeting times, but still requires oversight. this shift moves beyond older gpt-3.5 implementations to offer higher intelligence, signaling a trend where ai agents handle increasing amounts of routine communication.