Robin Rombach
Showing 1–3 of 3 transcripts.
AMD, Starcloud, Coatue..10 Hot Takes From The Biggest Names in AI
Max Cook, Matan Grinberg, Pim de Witte, Philip Johnston, Mark Papermaster, Rodrigo Liang, Ramin Hasani, Robin Rombach, Chris Madden, Henri Delahaye, Molly, Philippe, Thomas
At the RAISE Summit, industry leaders like Max Cook and Matan outlined a decisive shift toward agentic AI ecosystems that replace traditional interfaces with natural language and naturalize the routing between frontier and open-source models based on latency and cost. Concurrently, hardware giants such as SambaNova and StarCloud announced billion-dollar investments and space-based compute strategies to overcome power constraints and enable efficient inference on edge devices. This convergence of sovereign AI initiatives, supply chain scaling, and automated software engineering marks a global transition where enterprises treat AI as a deployed product rather than a research tool.
- All-In Podcast1h 4m
Open Source Wins, AGI Is Here, and Scorsese’s AI Toolkit with CEOs of Cerebras & Black Forest Labs
Andrew Feldman, Robin Rombach, Martin Scorsese
Cerebras Systems leverages a $25 billion order backlog from major tech giants to deploy chips that accelerate AI reasoning fifteen times faster than standard inference, effectively achieving early Artificial General Intelligence while reshaping global energy infrastructure. Simultaneously, Black Forest Labs utilizes its pioneering latent diffusion algorithms to bridge generative media and robotics, partnering with entities like Martin Scorsese to create scalable production tools and autonomous "world models." Both companies advocate for a hybrid ecosystem combining open-source accessibility with proprietary sovereignty, aiming to resolve economic bottlenecks and regulatory challenges through staged deployments and secure data practices.
- a16z39 min
Text to Video: The Next Leap in AI Generation
Anjney Midha, Andreas Blattmann, Robin Rombach
Released on November 21st, Stable Video Diffusion is a state-of-the-art open-source model that generates short video clips from single input images by prioritizing the learning of complex physical properties like 3D consistency and camera movement. The architecture employs diffusion methodology over autoregressive methods to optimize perceptual details and utilizes LoRA adapters for scalable control of camera motion, while training strategies focused on temporal dynamics and specific 3D orbit refinement to achieve surprising reasoning capabilities in as few as 2,000 iterations. This release continues the team's philosophy of driving innovation through algorithmic efficiency rather than sheer compute volume, having already catalyzed a rapid ecosystem of community experimentation and set a roadmap for longer sequences and future audio integration.