⌘ Tech
Flower AI Day: Collaborative Superintelligence

- When
- Wednesday, September 30 · 1:00 PM – 8:30 PM
- Where
- The Pearl — Dogpatch
- Listed by
- Lu.ma — Gen AI SF
Flower AI Day returns this year as AI enters a new phase. The frontier of AI is no longer defined by single, isolated, closed models. Increasingly, it is being pushed forward by open agents and models working together, combining their capabilities and improving collectively. We call this collaborative superintelligence, and it will be the focus of Flower AI Day 2026.
Expect a world-class afternoon and evening program featuring some of the top thinkers and builders in AI. Past Flower events have included Sara Hooker (CEO, Adaption), Joseph Spisak (VP, Reflection), Ce Zhang (CTO, Together AI), Jonathan Frankle (Chief AI Scientist, Databricks), and Thomas Wolf (CSO, Hugging Face).
Join us for technical talks, new Flower technology and demos, expert firesides and panels, and plenty of opportunities to meet others building open, collaborative, and production-ready AI systems.
The Dawn of Collaborative Superintelligence
Flower AI Day will explore the technologies defining the emerging frontier of collaborative superintelligence, from cutting-edge research to systems already being deployed by ambitious organizations and enterprises:
Agent teams for long-horizon complex workOpen-source AI especially post-trained base models exceeding closed frontier AI systemsAI Grids and forms of distributed inference that maximize networked AI resourcesMulti-agent systems combining specialist capabilities and guided by user supervisionLocal AI, supporting on-device and private frontier-level open-weight LLMsDecentralized training for foundation models using multi-site compute and dataFederated learning and analytics methods targeting sensitive data domainsSovereign, compliant, and privacy-preserving model lifecyclesAI requiring confidential compute, attestation, governance, and policy enforcementReal-world enterprise AI deployments in healthcare, automotive, teleco, finance, and manufacturing
Complete details of the 2026 program, including speakers, will be announced soon.
Flower AI Day 2026
Our program kicks off early in the afternoon on Wednesday, September 30, bringing together the Flower community for an afternoon and evening of technical talks, demonstrations, discussion and networking.
We are delighted to be hosting Flower AI Day at The Pearl in San Francisco’s Dogpatch neighborhood, at the heart of a fast-growing AI and startup district around Y Combinator, OpenAI and Mission Bay. All attendees will receive our world-famous Flower swag, and the program will include several catered social and networking sessions. The day will conclude with an evening reception featuring a buffet dinner, open bar, complimentary cocktails and drinks.
Last year at Flower AI Day 2025, we brought together speakers and panelists from Cohere, Meta, Together AI, Red Hat, J.P. Morgan, AWS, Coinbase, Vana and Flower Labs. Speakers included Sara Hooker, Joseph Spisak, Ce Zhang, Tony James, Sudhir Upadhyay, Anna Kazlauskas, Erran Li, Kevin Leffew, Daniel J. Beutel, Nic Lane, Patrick Foley and others. We anticipate Flower AI Day 2026 will build on that momentum, bringing together an exceptional group of researchers, founders, engineers and industry leaders shaping the future of collaborative AI.
Preliminary Agenda
1:00 PM — ☕️ Registration and Reception
Check in for badge and swag. Network over coffee.
1:30 PM — 🌼 Welcome and Flower Keynote
A first look at brand new Flower tech for collaborative superintelligence, with demos of open agents and models that tackle frontier level AI tasks including long-horizon work.
2:10 PM — 🎙️ Building the Open Frontier Together
From 80B-scale model training to open models that organizations can build on and improve together. Explore how decentralized training and collaboration across institutions can put frontier capability into more hands.
2:30 PM — Fireside Chat: Open Source and the Race to Superintelligence
3:00 PM — Refreshment Break and Networking
3:20 PM — 🎙️ Scaling Enterprise AI Without Giving Up Control
A…
