My top 5 takeaways on AI workflows @akothari and @_ryannystrom (Notion's co-founder and AI engineering lead): 1. Focus AI on automating busywork The turning point was “when we decided to stop building toy apps that look like AI slop and instead focus on actually helping you get your work done.” Don’t chase shiny AI trends—solve the annoying manual tasks users hate doing. 2. Assign work to AI agents using your voice From Akshay: “I’ve completely moved away from writing a doc or filling a database manually to just talk to the agent like 90% of the time.” Akshay lets AI handle tedious work by just sharing what he wants with his voice. 3. Structure your data the way AI models like to consume it Notion’s AI agent breakthrough came when the team stopped trying to get the AI model to understand their Notion page structures and instead formatted everything as markdown (a format that AI already understands). 4. Show the AI working in real-time to build trust Ryan designed Notion's agent so “I can see it write in real time. I can see it fill in a database in real time.” When users can watch the agent work step-by-step instead of waiting for a black-box result, they trust it more and know when to intervene. 5. Make it easy for power users to teach others “Out of a thousand people, there might be one person who’s just cracked the code of how to use your AI product. Now that one person can impact a thousand people because of something they built.”