This week, we’re joined by Dr. Luis Serrano of the Serrano Academy, an expert in making Math and Quantum Machine Learning accessible. We discuss simplifying complex A.I. topics, Cohere’s innovative use of embeddings, and the future of Quantum ML.
About Luis Serrano
Luis Serrano is the creator of the popular YouTube channel Serrano.Academy, and the author of the Amazon Bestseller Grokking Machine Learning. Luis has a PhD in mathematics from the University of Michigan, and spent some years as a researcher in academia. He then made a switch to technology, and worked in AI and quantum computing at companies like Google, Apple, Cohere, and Zapata Computing, and has popular machine learning courses in platforms such as Coursera and Udacity. Luis is a lifelong educator, and he strives to simplify complex concepts and make education accessible to all.
Overview
Dr. Luis Serrano pulls back the “curtain of abstraction” to reveal that AI and math aren’t inherently difficult but are made to seem so. He believes people struggle with math not because they are bad at it, but because they struggle with abstract concepts. Reflecting on his own experience, Luis shares how he once disliked math until he won a national math competition that focused on puzzles. This revealed his love for puzzles and sparked his passion for mathematics, leading him to become a mathematician.
As the founder of the Serrano Academy, which has over 140,000 followers, Luis acknowledges that creating educational content can take a long time, with some videos taking years to complete. During his time at Udacity, he learned that people digest information in three main ways: theoretically, visually, and through hands-on building. By addressing all three types of learners, Luis makes his content accessible to a broader audience.
Luis recently worked as the Head of Developer Relations at Cohere, a leading lab in LLMs. He explained Cohere’s focus on enterprise tools, especially embeddings. According to Luis, improved embeddings have significantly enhanced LLMs, making complex problems easier to solve. He highlighted Cohere’s Rerank, which boosts the quality of keyword and vector searches without needing a complete system overhaul.
Looking ahead, Luis discusses the future of AI beyond generative chat, including multimodality, intelligent agents, and real-world interactions with robots and autonomous vehicles. He also talks about promising applications for quantum machine learning. As quantum computers increase their capacity with more qubits, exciting advancements in quantum ML are expected.
In this episode you will learn:
- How math and AI can be made easy to understand [05:21]
- The three major categories of learners [16:21]
- Why embeddings are the most important component of LLMs [26:19]
- How semantic search differs from a traditional keyword search [29:57]
- The most exciting emerging application areas for AI [42:41]
- The promising application areas for Quantum Machine Learning [49:18]
Items mentioned in this podcast:
- This episode is brought to you by AWS Trainium and AWS Inferentia
- Serrano Academy
- Serrano Academy YouTube channel
- Grokking Machine Learning by Luis Serrano
- Udacity
- Cohere
- DeepLearning.AI
- Luis’ course with DeepLearning.AI
- Cohere Rerank
- Zapata Computing