Yuri Plotkin, a Biomedical Engineer and Machine Learning Scientist, dives into his journey from biology to AI, driven by curiosity. He discusses generative AI and diffusion models, tracing their evolution and potential across industries. Highlighting the intricate relationships between various machine learning models, he uses analogies and humor to illustrate concepts. Yuri emphasizes the need for a blend of theory and practice in machine learning engineering, while addressing the complexities of deploying AI in diverse sectors.
Read more
AI Summary
AI Chapters
Episode notes
auto_awesome
Podcast summary created with Snipd AI
Quick takeaways
Yuri Plotkin emphasizes the historical interconnectedness of generative AI models, underscoring the importance of foundational knowledge in driving advancements.
The Variational Book aims to clarify complex generative AI concepts for students, focusing on Bayesian latent models while intentionally excluding LLMs.
Deep dives
The Evolution of AI Models
The conversation highlights the significant lineage of generative AI models, emphasizing how each innovation builds on its predecessors. By tracing the historical development of these models, it becomes evident how foundational technologies inspire more advanced versions, such as the journey from traditional image generators to cutting-edge text-to-video applications. This progression reveals the importance of understanding prior innovations in order to fully grasp contemporary techniques, reaffirming the saying about 'standing on the shoulders of giants.' This exploration emphasizes the interconnectedness of various generative AI models, showcasing how the accumulation of knowledge propels the field forward.
Inspiration for The Variational Book
The author reflects on the motivation behind writing 'The Variational Book,' stemming from a desire to synthesize and clarify complex information about generative AI algorithms. As he engaged with numerous academic papers, he noticed common threads that illustrated the foundational principles behind the various models. The book aims to present a step-by-step logical framework, breaking down these concepts for readers, including undergraduate and graduate students in the field. The intent is also to make complex subjects more accessible, eliminating gaps in understanding while enriching the reader's knowledge.
Excluding LLMs for Focus
A notable design choice in the book is the exclusion of large language models (LLMs), allowing the author to concentrate on Bayesian latent models and variational inference methods instead. This decision was influenced by the vast amount of existing literature on LLMs already available, indicating that focusing on less-covered areas could provide greater value. The author believes that the theoretical foundations and methodologies in his book still apply to modern techniques, enhancing understanding across the field of AI. This focus on traditional models underlines the significance of solidifying foundational knowledge to inform future developments in technology.
Efficiency and Practicality in Video Generation
The discussion touches on the practical implications of generative model advancements, particularly in the realm of video generation. The evolution of diffusion models shows a trend toward reducing the number of steps necessary for generating high-quality results, which directly correlates to more timely video output. As generative AI technologies progress, faster inference processes facilitate the creation of longer videos, which has wider applicability across various industries, including entertainment. This intersection of efficiency and innovation indicates a promising future for generative AI in transforming how visual content is produced.
The Variational Book // MLOps Podcast #253 with Yuri Plotkin, an ML Scientist.
// Abstract
Curiosity has been the underlying thread in Yuri's life and interests. With the explosion of Generative AI, Yuri was fascinated by the topic and decided he needed to learn more. Yuri pursued learning by reading, deriving, and understanding seminal papers within the last generation. The endeavors culminated in the writing of a book on the topic, The Variational Book, which Yuri expects to release shortly in the coming months. A bit of detail about the topics he covers can be found here: www.thevariationalbook.com.
// Bio
Evolved from biomedical engineer to wet-lab scientist, and more recently transitioned Yuri's career to computer science with the last 10+ years developing projects at the intersection of medicine, life sciences and machine learning.
Yuri's educational background is in Biomedical Engineering, at Columbia University (M.S.) and University of California, San Diego (B.S.). Current interests include generative AI, diffusion models, and LLMs.
// MLOps Jobs board
https://mlops.pallet.xyz/jobs
// MLOps Swag/Merch
https://mlops-community.myshopify.com/
// Related Links
Website: https://plotkiny.github.io/
The Variational Book: www.thevariationalbook.com
SAS: https://www.sas.com/en_us/home.html
SAS® Decision Builder: https://www.sas.com/en_us/offers/23q4/microsoft-fabric.html