"Sounds good!" "Thanks!" "Got it!" Three generic buttons that work 2% of the time.
Gmail gives you autocomplete. BadAtMail gives you an actual assistant who reads context, remembers what matters, and handles your email like a real person would.
Try BadAtMailSmart Reply generates three-word responses based on the last message. It doesn't understand context, tone, or what you actually care about. It's pattern matching, not comprehension.
The AI reads every email in your inbox. Understands threads and knows what's urgent. Remembers your preferences, learns what you care about. Then drafts full replies that actually address what was asked.
You still have to read the email and decide what to do with it. Write the reply if it's not generic enough for "Thanks!" You're still doing all the cognitive work.
AI classifies incoming emails. Archives the stuff you don't care about. Drafts replies for things that need responses. Only bothers you with what actually matters. You just review and approve.
"Let me know if you have any questions!" as a reply to someone asking you multiple specific questions. We've all seen it suggest something hilariously inappropriate because it doesn't actually understand the conversation.
The AI learns your writing style, tone, and voice. It can generate multi-paragraph responses that address every point. It references previous context. It sounds like an actual human wrote it - specifically, like you wrote it.
Gmail's Smart Reply is a feature. A small enhancement to make clicking buttons slightly easier. BadAtMail is a fundamental rethinking of how email should work - treating it as an automation problem, not a UI problem.