Reimagining the email experience with AI

Superhuman has spent years selling speed: keyboard shortcuts, instant search, an inbox that feels like it's daring you to hit zero. Now the company is betting that speed alone isn't the point anymore, and it's leaning on OpenAI to prove it. The pitch, laid out in a joint post from OpenAI, is that email should stop being a chore you manage and start being a task an AI handles alongside you.

What that means in practice is less about faster typing and more about judgment. Instead of a smarter autocomplete bolted onto an old interface, Superhuman is folding OpenAI's models into the core of how the app reads, triages, and responds to messages. The idea is that the software should understand context well enough to draft a real reply, not just finish your sentence, and know which threads actually need your attention versus which ones can wait or vanish into an archive.

This fits a pattern we've seen accelerate over the past year: productivity software rebuilding itself from the model outward rather than sprinkling AI features on top of legacy tools. Notion, Superhuman, and a growing list of workplace apps are treating GPT-class models less like a chatbot widget and more like an engine sitting underneath the entire product. For email specifically, that's a meaningful shift, because inbox management has resisted automation for two decades despite endless startups promising to fix it.

Whether users actually want an AI reading and drafting their email, rather than just filtering spam, is the open question here. Superhuman's audience skews toward people already obsessed with inbox efficiency, so it's a reasonable testing ground. But the real test comes when this kind of AI-first email design reaches people who don't care about keyboard shortcuts and just want their inbox to stop being annoying.