Open Source Causal AI & The Generative Revolution | Emre Kıcıman Ep 16 | CausalBanditsPodcast.com
May 20, 2024
auto_awesome
Emre Kıcıman, a core developer of DoWhy, delves into open source causal AI with Microsoft and Amazon collaboration. They discuss the core of science in causal AI, the intersection of language models and world models, the usefulness of modeling physics, and the future of generative AI. The conversation explores challenges in causality, the importance of causal inference in decision-making, and the evolution of libraries promoting causal analysis.
Large language models can aid causal analysis by assisting in setting up causal assumptions and critiquing them.
Generative models offer the potential to approximate causal world models, enhancing artificial intelligence advancements.
Deep dives
Causality in Large Language Models
Emre Kijeman discusses the potential of large language models to contribute to causal analysis. While current models may not reason causally, their embedded knowledge offers opportunities to enhance causal analysis by aiding in setting up causal assumptions and critiquing them. This augmentation complements existing methods and provides domain experts with valuable support in identifying plausible causal mechanisms.
Utilizing Large Language Models for Knowledge Graph Construction
The podcast explores the use of large language models in constructing knowledge graphs alongside domain experts. By leveraging LLMs, organizations like BMW Group have accelerated the knowledge graph construction process, motivating experts to contribute their valuable insights. The models not only enhance efficiency but also inspire greater engagement and contributions from domain experts by presenting initial graphs for further refinement and validation.
Evolution of Causal Analysis in AI
The discussion delves into the transformative impact of generative models on artificial intelligence, emphasizing the importance of incorporating causal reasoning in modeling the world. While current models may not explicitly learn causal relationships, the potential exists for them to approximate causal world models. The future direction involves exploring how these models can advance by gaining true causal understanding and volumetric insights.
Future Directions in Causal Analysis
Emre Kijeman shares insights on furthering causal analysis by integrating foundation models to support more complex physics-style systems. The aim is to enhance the practical use of large language models in suggesting causal graphs, critiquing analyses, and identifying missing confounders. Additionally, efforts are directed towards utilizing LLMs to empower causal analysis in diverse and impactful applications.
What makes two tech giants collaborate on an open source causal AI package?
Emre's adventure with causal inference and causal AI has started before it was trendy.
He's one of the original core developers of DoWhy - one of the most popular and powerful Python libraries for causal inference - and a researcher focused on the intersection of causal inference, causal discovery, generative modeling and social impact.
His unique perspective, inspired by his experience with low-level programming combined with his vivid interest in how humans interact with technology, is driven by a deep seated desire to solve problems that matter to people.
In the episode we discuss: 🔹 What makes Microsoft and Amazon collaborate on an open source Python package? 🔹 Causal AI and the core of science 🔹 Is language model a world model? 🔹 When modeling physics is useful?
Ready to dive in?
Join the insightful discussions at https://causalbanditspodcast.com/
About The Guest Emre Kıcıman, PhD is a Senior Principal Research Manager at Microsoft Research. He's one of the core developers of the DoWhy Python package, alongside Amit Sharma. He holds a PhD in computer science from Stanford University. Privately, he loves to climb and spend time with his family.