Keynotes, interviews, podcasts and panels — mostly on AI research, open models, Tech’s opportunity, and what it takes to ship at billion scale. Click a row to play or read.
| 2026 | ||
| 2026 | ▶ | Machines Can Think · Abu Dhabi |
Industrial Day keynote at the UAE’s flagship AI adopters summit, co-hosted by Polynome and MBZUAI at Park Hyatt Saadiyat — on what comes after human-level perception, and what machines that see and act mean for industry. | ||
| 2025 | ||
| 2025 | ▶ | Meta · LlamaCon |
Meta’s first AI developer event. Opening keynote with CPO Chris Cox and research scientist Angela Fan — Llama 4, the developer ecosystem, and what comes next for open models. | ||
| 2024 | ||
| 2024 | ▶ | Indiaspora AI Summit |
On Llama 3, what open-weights at frontier scale unlocks, and how Meta thinks about the road to general intelligence. | ||
| 2024 | ◆ | SXSW Sydney |
A featured session with Ryan Patel on whether AI will be governed by a handful of closed players or by an open, inclusive ecosystem — and what that choice means for safety, innovation, and global access. | ||
| 2024 | ◆ | Fortune Brainstorm AI |
In conversation with Fortune’s Jason Del Rey at Brainstorm AI — on the false binary between open and closed AI, the case Llama makes for open-source development at the frontier, and where Meta goes as it builds toward AGI. | ||
| 2024 | ✎ | Capital Brief |
On power as the binding constraint for AI, why compute will spread regionally, and what that means for countries outside the US. | ||
| 2021 | ||
| 2021 | ◉ | How AI Happens · Sama |
A long-form conversation on the state of computer vision, the shift from third-person to first-person sensing, and the privacy & transparency tradeoffs in egocentric perception. | ||
| 2021 | ▶ | ICCV 2021 Workshop |
For the ICCV ‘21 “Share Stories and Lessons Learned” workshop — the academic and corporate journey, and what good research practice looks like across both. | ||
| 2021 | ▶ | KAUST Research Open Week |
Keynote on the perception & robotics agenda inside Meta AI Research — pushing the frontier of computer vision, conversational AI, and multi-modal reasoning. | ||
| Earlier | ||
| 2020 | ▶ | Invited keynote |
A keynote on the state of computer vision at scale and the platform problems that come with it. | ||
| 2017 | ▶ | LDV Vision Summit |
From LDV Capital’s 4th Annual Vision Summit. Inside the visual-understanding stack behind hundreds of Facebook products — the platform that became Lumos. | ||
| 2017 | ▶ | AI Frontiers Conference |
Video is becoming ubiquitous — from capture to consumption. If we can make machines understand video the way humans do, a long set of applications opens up. A walk through the obvious problems (datasets, action recognition, multi-modal understanding, temporal aggregation, appearance and motion, compression) and the longer-term directions: self-serve content understanding, large label embedding, video summarization. | ||
| 2017 | ▶ | Video @Scale · Facebook |
At Video @Scale — on the challenges of understanding video at Facebook’s scale, the recent efforts of the team, and a few exciting results from the platform that became Lumos. | ||
| 2016 | ▶ | Skolkovo Robotics · Moscow |
At Skolkovo Robotics in Moscow — on the computer vision work happening at Facebook scale, leading FAIR’s computer vision group. | ||
Speaking & press
Open to keynotes, panels, podcasts and interviews — especially on topics in multimodal, frontier models, AI platforms and what it takes to OSS models and build products for billion scale.
For booking, please include event, date, audience size and format.