
Hosted by Jon Krohn · EN

Jazmia Henry joins Jon Krohn to break down what it actually takes to build end-to-end foundation models for the energy industry. From wrangling decades of handwritten oil-and-gas documents into usable training data, to bespoke tokenizers, reinforcement learning, and inference at scale, Jazmia walks through every stage of the stack. Along the way she explains why reinforcement learning models are "bursty," what reward hacking is and how her Grounded Continuous Evaluation framework fixes it, and revisits the 2023 NeurIPS paper that argued, to widespread skepticism at the time, that scaling bad data degrades model performance. Additional materials: https://www.superdatascience.com/995 Interested in sponsoring a SuperDataScience Podcast episode? Email natalie@superdatascience.com for sponsorship information. In this episode you will learn: (10:06) The User Agnosticism Tenet (20:02) The Zillow Offers parable (23:25) Why workflows should come before agents (29:57) Why data engineering is the bedrock of AI (52:41) Why velocity is the only durable moat

Unemployment for recent computer-science graduates now rivals rates for fine-arts and anthropology majors, and undergraduate CS enrollment fell 11% in 2025. In this Five-Minute Friday, Jon Krohn walks through the data on both sides of the debate, from Stanford research showing a 13% employment drop for young workers in AI-exposed jobs, to Federal Reserve studies finding no statistically detectable link between AI adoption and reduced hiring. Jon shares his own view on where the truth lies and offers five concrete pieces of advice for graduates and senior professionals alike on how to get hired in 2026. Additional materials: www.superdatascience.com/993 Interested in sponsoring a SuperDataScience Podcast episode? Email natalie@superdatascience.com for sponsorship information.

For years, AI content has come in the form of “use this library, use this tool” tutorials that age out within months. Jacob Miller and Jeremy Mumford, co-authors of the brand new Wiley book Architected Intelligence, wanted to write something different, a guide to the higher-level principles of building AI products and AI-first organizations that will still be relevant in five or ten years. In this episode, the two Pattern engineers walk Jon Krohn through the core ideas of their book: why you should design products and processes so they can be executed by a human, an AI agent, or any hybrid combination; why most companies are still treating hallucinations as a model problem when they’re actually a data curation problem; why the natural progression of AI development goes skills, workflows, agents, not straight to agents; and why velocity, not models or data, is the only durable competitive advantage left. Additional materials: https://www.superdatascience.com/993 Interested in sponsoring a SuperDataScience Podcast episode? Email natalie@superdatascience.com for sponsorship information. In this episode you will learn: (10:06) The User Agnosticism Tenet (20:02) The Zillow Offers parable (23:25) Why workflows should come before agents (29:57) Why data engineering is the bedrock of AI (52:41) Why velocity is the only durable moat

While “tokenmaxxing”, the social media trend of maximizing AI token consumption as a vanity metric, takes off online, the physical infrastructure behind AI is slamming into serious bottlenecks. In this Five-Minute Friday, Jon Krohn maps out the four overlapping supply-chain constraints choking AI compute: GPUs (with NVIDIA Blackwell sold out through mid-2026), high-bandwidth memory (quintupled demand since 2023, only three manufacturers worldwide), CPUs (agentic AI requires 12x more CPUs per GPU than chatbots), and electricity (Gartner projects power shortages will restrict 40% of AI data centres by 2027). Find out why the five biggest hyperscalers are on track to spend $725 billion on AI infrastructure in 2026, where the reasons for optimism lie, and why Jon says you should definitely not tokenmaxx. Additional materials: www.superdatascience.com/992 Interested in sponsoring a SuperDataScience Podcast episode? Email natalie@superdatascience.com for sponsorship information.

Dr. Trevor Manz of Marimo talks to Jon Krohn about Marimo Pair, an open-source agent skill that teaches coding agents like Claude Code how to drive a reactive Python notebook, reading cell state, running Python in the kernel, taking screenshots of cells, and iterating on data tasks the way agents iterate on traditional software. Trevor also unpacks recursive language models, his AnyWidget project that bridges Python and the web, and his journey from a Wisconsin small town and Harvard bioinformatics research to founding-engineer life at Marimo. Listen to the episode to hear why no matter where AI takes us, curiosity and going deep on a topic will always be valuable. Additional materials: www.superdatascience.com/991 Interested in sponsoring a SuperDataScience Podcast episode? Email natalie@superdatascience.com for sponsorship information. In this episode you will learn: (07:04) What Marimo Pair is and how it teaches agents to use notebooks as a tool (13:03) How agent skills work as folders of markdown files (24:15) Trevor's day-to-day workflow combining Claude Code and Marimo Pair (31:51) Recursive language models and why they could be the future of agentic reasoning (57:33) Career advice on curiosity, going deep, and becoming a domain expert

Anthropic has built a frontier AI model so capable at finding software vulnerabilities that it has decided not to release it to the general public. In this Five-Minute Friday, Jon Krohn breaks down Claude Mythos Preview, a general-purpose model whose hacking abilities emerged as a side effect of broad improvements in code understanding and reasoning. Find out how Mythos achieved a nearly 100x improvement over Opus 4.6 on Firefox exploit generation, why Mozilla patched 271 vulnerabilities in a single release using an early version of the model, and what Project Glasswing Anthropic’s gated industry consortium means for the future of cybersecurity. Jon also shares practical tips for securing the code you’re generating with AI tools. Additional materials: www.superdatascience.com/990 Interested in sponsoring a SuperDataScience Podcast episode? Email natalie@superdatascience.com for sponsorship information.

Rubrik’s Anneka Gupta and Cal Al-Dhubaib speak to Jon Krohn about cybersecurity measures, the risks AI in business might pose for malicious attacks, and why AI should be kept “boring.” Find out how Rubrik safeguards client data, what zero trust is in the context of cybersecurity, and why cyber-resilience needs to be a top priority for companies looking to adopt AI. Additional materials: www.superdatascience.com/989 Interested in sponsoring a SuperDataScience Podcast episode? Email natalie@superdatascience.com for sponsorship information. In this episode you will learn: (02:25) All about Rubrik (08:51) The announcement of Claude Mythos (26:26) Utilizing zero trust (40:36) About the Rubrik agent cloud

In this month’s episode of In Case You Missed It, Jon Krohn talks to guests about memory and education, and how artificial intelligence is continuing to help lower the barriers to access. Hear from Matt Glickman, Traci Walker-Griffith, Richmond Alake, and Linda Haviv, discussing the foundations of AI agent memory, how engineers can develop at scale, and why they believe AI could be your child’s perfect tutor in the classroom. Additional materials: www.superdatascience.com/988 Interested in sponsoring a SuperDataScience Podcast episode? Email natalie@superdatascience.com for sponsorship information.

Linda Haviv talks to Jon Krohn about staying current on AI matters, why open-source technology is narrowing the gap in its race with proprietary models, and how being a content creator in tech is key to career growth and longevity. She emphasizes that non-linear pathways to a career in tech can give applicants an edge, and stresses the importance of continuous upskilling to “stay relevant.” In her view, systems thinking is becoming more important than coding skills. Hear why in this episode. Additional materials: www.superdatascience.com/987 Interested in sponsoring a SuperDataScience Podcast episode? Email natalie@superdatascience.com for sponsorship information. In this episode you will learn: (03:43) Linda Haviv on AI education (13:16) The future of coding (27:00) Having a side hustle in today’s economy (31:01) On becoming a content creator for tech (1:00:14) How open source could disrupt the AI landscape

CTO of Propel Software Kishore Subramanian talks to Jon Krohn about how product lifecycle management (PLM) software and quality management systems (QMS) help ensure compliance, record management, and quality assurance. Listen to the episode to hear Kishore Subramanian talk about best practices for getting started with Agentforce 360, his top tips for deploying AI projects, and why yoga and meditation could make you better at building AI products! Additional materials: www.superdatascience.com/984 Interested in sponsoring a SuperDataScience Podcast episode? Email natalie@superdatascience.com for sponsorship information. In this episode you will learn: (05:21) How Propel Software meets its customers’ demands (07:57) About Propel One AI (13:31) A case study for Salesforce’s Agentforce 360 Platform (17:08) How to build an enterprise-ready agent with Agentforce 360 (19:21) How to get your AI tool into production