Breaking News: Major Collaboration to Scale Reinforcement Learning
NVIDIA and Ineffable Intelligence, the London-based AI lab founded by AlphaGo architect David Silver, today announced an engineering-level collaboration to build the infrastructure needed for large-scale reinforcement learning (RL). This partnership aims to create systems that learn by trial and error, converting computation into new knowledge—marking a shift beyond AI models that rely solely on human data.
"The next frontier of AI is superlearners—systems that learn continuously from experience," said Jensen Huang, founder and CEO of NVIDIA. "We are thrilled to partner with Ineffable Intelligence to codesign the infrastructure for large-scale reinforcement learning as they push the frontier of AI."
David Silver, a pioneer in reinforcement learning, highlighted the paradigm shift: "Researchers have largely solved the easier problem of AI: building systems that know all the things humans already know. But now we need to solve the harder problem: building systems that discover new knowledge for themselves."
Technical Challenge: Data on the Fly
Unlike pretraining with fixed datasets, reinforcement learning systems generate their own data during operation. The system must act, observe, score, and update in tight loops, placing unprecedented pressure on interconnect, memory bandwidth, and serving infrastructure. Engineers from both companies are working to build a pipeline that can feed these systems at scale, starting with NVIDIA Grace Blackwell and extending to the upcoming Vera Rubin platform.
This approach requires novel model architectures and training algorithms, as the systems will learn from rich forms of experience distinct from human language. The goal is to understand the hardware and software needed as AI moves beyond human data toward simulation-based learning.
Background: Ineffable Intelligence Emerges from Stealth
Ineffable Intelligence emerged from stealth mode just last week, revealing its focus on reinforcement learning and its founder, David Silver—widely recognized for his work on AlphaGo. The lab's mission aligns with NVIDIA's long-standing investment in AI infrastructure. The collaboration builds on years of research into RL, which has already transformed areas like robotics, game playing, and autonomous systems.
What This Means: Unlocking Discovery Across Fields
Getting this infrastructure right could unlock unprecedented scale for reinforcement learning in highly complex environments. Agents trained in simulation could discover breakthroughs in science, medicine, engineering, and beyond—generating knowledge autonomously. NVIDIA and Ineffable are betting that this paradigm shift will define the next era of artificial intelligence, moving beyond imitation to true discovery.
As Silver noted, "That requires a very different approach—systems that learn from experience." With this partnership, the industry takes a concrete step toward making that vision a reality.