🔬Searching the Space of All Possible Materials — Prof. Max Welling, CuspAI - Latent Space: The AI Engineer Podcast Recap

Podcast: Latent Space: The AI Engineer Podcast

Published: 2026-02-25

Duration: 34 min

Summary

In this episode, Prof. Max Welling discusses the intersection of physics and AI, emphasizing the need for a new approach to material science via his startup CuspAI. He reflects on his career trajectory and the growing importance of making an impact through technology in the face of climate change.

What Happened

The episode opens with a fascinating analogy by Prof. Max Welling, who likens nature to a 'physics processing unit' (PPU), suggesting that nature itself can perform computations when it comes to understanding new materials. He elaborates on the challenges of programming nature for experiments, emphasizing the potential for combining digital computations with natural processes to discover novel materials. This perspective sets the stage for Welling's broader vision of how AI can interface with the physical world to solve pressing scientific problems.

Welling shares insights from his extensive career, highlighting his contributions to deep learning and material science. He explains how his focus has evolved from purely theoretical physics to prioritizing impactful work, particularly in response to the climate crisis. The conversation delves into his motivations for founding CuspAI, where he aims to leverage AI tools to address fundamental questions in material science while also creating tangible benefits for society. Welling articulates a duality in his work: chasing deep scientific questions while aiming for real-world impact, particularly in the context of environmental sustainability.

The discussion shifts to the emerging discipline of AI for science, which Welling describes as an exploding field with significant investment momentum. He notes that advancements in AI tools have made their application in scientific research more feasible and effective. Welling’s reflections on this trend underscore a broader shift in the scientific community towards integrating AI methodologies to tackle complex challenges, thereby revolutionizing traditional research paradigms.

Key Insights

• Nature can be viewed as a physics processing unit, performing computations that can aid material discovery.
• Welling's career trajectory reflects a shift towards making a positive impact on society through technology, particularly concerning climate change.
• The intersection of physics and AI is creating new opportunities in material science and beyond.
• AI for science is rapidly gaining traction, with significant investments signaling a transformative shift in how research is conducted.

Key Questions Answered

What is the physics processing unit concept discussed by Max Welling?

Prof. Welling introduces the idea of a physics processing unit (PPU), suggesting that nature itself can compute solutions to scientific problems. He contrasts traditional digital processing with this notion, illustrating how nature may perform computations that lead to new material discoveries. This concept emphasizes the complexity of interfacing with natural processes while highlighting the potential synergy between computational methods and the physical world.

How did Max Welling's career focus evolve over time?

Welling reflects on his career trajectory, noting that his focus has transitioned from theoretical explorations in physics to addressing real-world problems with potential societal impact. He acknowledges that while early interests were driven by curiosity about fundamental scientific questions, a sense of urgency regarding climate change has compelled him to prioritize work that can yield tangible benefits. This evolution has culminated in the establishment of CuspAI, aimed at leveraging AI to tackle material science challenges.

What role does AI play in addressing climate change according to Welling?

Welling expresses concern over climate change and the slow pace of political solutions, advocating for a technological approach to the crisis. He believes that by harnessing AI, significant advances can be made in developing new materials that could contribute to sustainability efforts. His startup, CuspAI, embodies this vision, merging cutting-edge AI techniques with material science to generate innovative solutions that could positively affect the environment.

What is the significance of the emerging field of AI for science?

Prof. Welling highlights the rapid growth of the AI for science field, noting substantial investments that indicate a burgeoning area of research and application. He points out that as AI tools become more accessible, their integration into scientific methodologies is transforming how researchers approach complex problems. This shift is characterized by an explosion of interest and funding, revealing the potential for AI to revolutionize traditional scientific processes.

How does Welling connect physics to machine learning?

Welling elaborates on the deep connections between physics and machine learning, emphasizing the role of symmetries in both fields. He notes that his background in theoretical physics has equipped him with mathematical tools that are applicable to machine learning challenges. His work with graph neural networks and equivariance stems from this intersection, demonstrating how fundamental physical principles can inspire advancements in AI algorithms and applications.