Super Data Science: ML & AI Podcast with Jon Krohn
AI Infrastructure, Agentic Workflows, and Model Architecture with a focus on strategic insights and insider perspectives.
Nutrition Label
This channel delivers high-signal interviews with primary industry sources, offering deep strategic insights into AI infrastructure and model development. While transparency regarding guest affiliations is exceptional, the format relies on conversational claims rather than hands-on code demonstrations or independent benchmarks. Viewers get direct access to founder-level perspectives but should treat technical claims as expert testimony rather than verified tests.
Strengths
- +
- +
- +
Notes
- !Check video descriptions for disclosures, as the host often holds professional roles at the companies being discussed.
- !Discussions rely on guest claims and strategic frameworks rather than live coding or independent performance benchmarks.
Rating Breakdown
Breakdown across the key dimensions we rate. Methodology →
Recent Videos
Why Open Models Make Economic Sense for Startups (with Lin Qiao)
LLM Selection Has Become Exhausting (with Lin Qiao)
972: In Case You Missed It in February 2026 — with Jon Krohn @JonKrohnLearns
Why Enterprises Will Build Their Own Specialized AI Models (with Lin Qiao)
Autonomous Intelligence vs AGI (with Lin Qiao)
90% of The World’s Data is Private; Lin Qiao’s Fireworks AI is Unlocking It
The Limits of Creativity in Large Language Models (with Tom Griffiths)
Should AI Be Designed Like Human Intelligence? (with Tom Griffiths)
Why Probability Replaced Logic in AI (with Tom Griffiths)
The “100x Engineer”: How to Be One, But Should You? — with Jon Krohn (@JonKrohnLearns)
Are We Overestimating Neural Networks? (with Tom Griffiths)
Is There a Mathematical Theory of the Mind? (with Tom Griffiths)
The Laws of Thought: The Math of Minds and Machines, with Prof. Tom Griffiths
What Samsara Looks for When Hiring AI Engineers (with Praveen Murugesan)
When "Optimal" Algos Fail in the Real World (with Praveen Murugesan)
Why this rating
Evidence receipts showing why each dimension is rated the way it is.
“CEO of Lightning AI Will Falcon speaks to podcast host and Lightning AI fellow @JonKrohnLearns”[Description] →
The video description explicitly discloses the host's material connection (Fellow) to the company being interviewed, ensuring high transparency about the nature of the content.
“We use the accessibility tree... it's a much more compact representation of the DOM... but sometimes the accessibility tree is missing information, so we also use the screenshot.”[13:15] →
Demonstrates deep familiarity with the specific data structures (DOM vs. Accessibility Tree) and the edge cases encountered during training.
“How Will founded Lightning AI”[00:46:29] →
Provides a first-hand primary source account of the company's origin and the specific friction points that led to its creation.
“How Lightning AI Built a $500m ARR Full-Stack AI Cloud”[00:00] →
The title claims '$500m ARR' (Annual Recurring Revenue), but the content describes a merger with Voltage Park (which holds ~$500m in compute assets/capacity). Conflating potential capacity or asset value with realized ARR is a significant exaggeration.