A series of informal conversations with thought leaders, researchers, practitioners, and writers on a wide range of topics in technology, science, and of course big data, data science, artificial intelligence, and related applications. Anchored by Ben Lorica (@BigData), the Data Exchange also features a roundup of the most important stories from the worlds of data, machine learning and AI. Detailed show notes for each episode can be found on https://thedataexchange.media/ The Data Exchange podcast is a production of Gradient Flow [https://gradientflow.com/].
The podcast explores the landscape of AI hardware, particularly focusing on inference and the role of memory. Sid Sheth, founder and CEO of D-Matrix, discusses the limitations of SRAM and HBM for cloud inference, highlighting D-Matrix's focus on digital in-memory compute (DIMC) to address these challenges. Sheth explai...
The podcast explores the economics and challenges surrounding self-driving cars and the growing opposition to data centers. It highlights the shift in autonomous vehicle technology, comparing the sensor-heavy approach of Waymo to Tesla's camera-centric system, and debates the necessity of LiDAR for safety and public tr...
Ethan Ouyang, technical lead at Atoms, discusses why Chinese AI startups are launching in the U.S., citing the U.S. market's active building and monetization culture, willingness to pay for software, and a strong flywheel of capital, cloud infrastructure, and AI talent. He details Atoms' platform, designed to build rev...
The podcast explores the critical issue of security in the context of AI and AI agents, featuring Jason Martin, co-founder of Permiso Security. Martin discusses the increasing adoption of AI agents across various industries, from tech to insurance and casinos, noting that these agents, whether internal or external, pre...
The discussion centers on the evolution of data observability in the age of AI agents. Lior Gavish, CTO and co-founder of Monte Carlo Data, describes how data teams are increasingly building and deploying AI agents to automate tasks and enhance user experiences across various industries. A key shift involves the need f...
The podcast explores the challenges enterprises face in deploying AI models due to inherent risks like bias, hallucinations, and vulnerabilities, which often hinder ROI. Ben Lorica, CEO of HIRONDO, introduces machine unlearning as a solution to make AI trustworthy by enabling models to "forget" undesired data and behav...
The podcast episode analyzes key takeaways from the Consumer Electronics Show (CES) 2026, focusing on humanoids and software-defined vehicles. Humanoids were a major presence, though mostly tele-operated with limited real-world applications, raising questions about investor expectations and the impact on manufacturing ...
Genesis Computing's CEO Matthew Glickman discusses AI data agents and their application in automating data workflows, particularly for data engineers. He addresses the problem of AI projects getting stuck before deployment, emphasizing the need for reliability over mere accuracy. Glickman highlights the value of focusi...
The podcast features a panel discussion on the current state and future directions of agentic AI. The panelists share insights on what architectural patterns and tools are proving effective, with coding agents and the use of evals highlighted as key components for success. They explore the complexities of multi-agent s...
This podcast episode presents a curated list of nonfiction books from 2025, offering recommendations for holiday reading. The selections span technology, business, and global politics, with a particular emphasis on China's role in various industries. NVIDIA's rise in the AI sector and Masa Son's career in tech investin...
The podcast explores the unique security challenges presented by large language models (LLMs) and agentic AI, contrasting them with traditional software security. Steve Wilson, Chief Product Officer of Exabeam and OWASP contributor, highlights prompt injection, AI supply chain vulnerabilities, and sensitive information...
The podcast explores whether AI should be considered a utility by examining properties such as usability, accessibility, portability, and trust. Evangelos Simoudis argues that AI currently falls short, particularly in accessibility and the variability in output quality based on payment. He contrasts AI with established...
Kumo.ai's relational foundation model for enterprise data is the focus, with Jure Leskovec, Stanford professor and Kumo.ai co-founder, detailing its capabilities. Kumo.ai targets AI predictions and risk scoring over structured business data, offering an alternative to traditional data science teams or generic large lan...
Graph RAG and its progression from prototype to production are examined, focusing on how knowledge graphs enhance Retrieval Augmented Generation. Philip Rathle, CTO of Neo4j, explains that Graph RAG leverages the implied structure within unstructured data to improve AI application accuracy. He notes that while vector-b...
The podcast explores the interpretability of large language models, drawing parallels between their analysis and biological systems. Emmanuel Ameisen, an interpretability researcher, discusses how these models, unlike traditional programs, are "grown" through extensive training data, making their reasoning processes op...
The discussion centers on the impact of AI on layoffs and the necessity for companies to demonstrate a return on their AI investments. It is argued that AI is contributing to layoffs, particularly in junior roles and middle management. Companies are reorganizing, and AI is being used as a justification for some layoffs...
The podcast explores how data engineering can adopt software engineering principles to build more robust and scalable data lakehouses, especially with the rise of AI agents. Ciro Greco, co-founder and CEO of Bauplan, discusses how their platform enables data teams to treat data like code, using Git-like branching and v...
Tiger Data founders Mike Freedman and Ajay Kulkarni discuss their new "agentic Postgres," a database designed for AI agents, addressing its unique requirements compared to traditional databases. They argue that agents, unlike humans, require features like safe play environments, parallelization, and efficient context r...
The discussion centers on the growing importance of orchestration in AI and data pipelines, particularly context engineering. Nick Schrock, founder and CTO of Dagster, distinguishes between dynamic orchestration for agentic workflows and data pipeline orchestration, which Dagster focuses on. He emphasizes that a solid ...
Follow this podcast in Podwise
Sign in to get AI summaries, transcripts and mind maps for any episode, including new ones.
