Today we launched the new @loom for Mac by @Atlassian, rebuilt from the ground up as a native app https://www.atlassian.com/blog/loom/all-new-loom-for-mac It started with a question: if we were building Loom today, knowing what we know, what would it look like? and how would we do it "AI natively"? We were aiming for three things: faster, higher quality, and taking advantage of the rich experience native Mac apps have to offer. We wanted something so good people couldn’t ignore it, on a foundation that’s easier to build on. Our first bet: six weeks to build a proof of concept that could record a Loom. When a few leaders started playing with it, it was clear there was a lot more worth going after. AI let more of us turn ideas into things we could try together. Five things shaped what happened next. A thread:

Yesterday I taught a 3-hour Forge workshop at Team ’26 Europe in Amsterdam! We took meeting notes, used AI to understand the work inside them, created Jira tickets, preserved context in Confluence, added human approval, and handed eligible work to an agent. The thing I want software developers to understand is this: Forge is not just “build something inside Jira.” You can build real workflows across the Atlassian ecosystem with backend logic, state, queues, permissions, AI, and agents. If you’ve never built with Forge before, I hope that makes you think a little differently about what’s possible. I’m at the Forge booth today! Come build with me and geek out about what you can create!