⌘ Tech
AiOS Meetup - San Francisco Edition

- When
- Tuesday, October 20 · 6:00 PM – 8:30 PM
- Listed by
- Lu.ma — Convex
AiOS San Francisco is happening—and you’re invited! 🤖
Join us for an evening with people who are as excited about AI and Apple platforms as you are. Whether you’re building with the latest Apple frameworks, experimenting with on-device AI, or simply curious about what’s next, expect live demos, fresh ideas, and plenty of opportunities to connect.
Talks and demos
Off Script: Local AI on Apple Devices from an ex-Apple Guy by Todd DaileyTodd Dailey ran business messaging for Apple Intelligence and AI at Apple, after years briefing executives around the world. Now he runs pretty much every AI, everywhere he can. He'll show a couple of live demos and share stories from inside Apple, and where he thinks on-device AI is headed, and where it isn't.Managing the App Store in the age of AI by Jacob XiaoJacob was one of the creators of SwiftUI at Apple, then left to co-found Bitrig, an AI coding tool for Apple platforms. He'll show how Bitrig makes it easy to use AI to manage your App Store listing as code so you can review, version, and safely sync.Infinite Desktops Without Renting a Cloud by Francesco BonacciComputer use is becoming a core capability of AI agents like Hermes, OpenClaw, and Bots. But where should those agents actually run? Cloud VMs are expensive, while giving agents access to your own computer comes with risks. What if you could run dozens of isolated desktops on hardware you already own? This talk explores how agents can safely share computers with each other and with humans, including live supervision, secure access to signed-in apps, and lessons from early users. Plus, a live demo of Cua Spaces, an open-source solution that makes it possible.One LLM, Multiple Macs: Distributed Inference with Swift by Ronald MannakWhat if an AI model is too large for a single Mac? In this live demo, I'll show how to connect two Macs over Thunderbolt 5 and run a single LLM across both using MLX Swift, opening the door to running much larger models locally.
Partner
Convex