High Signal: Data Science | Career | AI

Steve Tadelis, Professor of Economics at UC Berkeley and former senior economist at eBay and Amazon, joins High Signal to bridge the gap between economic theory and the high-stakes reality of data science and AI. Drawing on his experience at the forefront of the world’s largest marketplaces, Steve discusses the "invisible friction" that prevents organizations from acting on data: a combination of misaligned incentives, organizational inertia, and the "Upton Sinclair problem," where leaders are effectively paid not to understand new paradigms.

The conversation moves from the "frustratingly obvious" opportunities left on the floor during eBay’s early years to the relentlessly scientific culture of Amazon. Steve explains why surface-level metrics like conversion rates often mask underlying rot in user retention and how rigorous experimentation, such as his famous $20 million search-ad experiment, can expose the difference between genuine growth and mere navigational intent. We also explore the structural shifts of the AI era, where Steve offers an important counter-narrative: rather than leveling the playing field, AI may act as an "unequalizer" that exponentially rewards those with the deepest critical thinking skills.

LINKS

Steve on LinkedIn

Consumer Heterogeneity and Paid Search Effectiveness by Blake, Nosko, and Tadelis (Econometrica, 2015)

The Limits of Reputation in Platform Markets by Nosko and Tadelis (NBER, 2015)

Information Disclosure as a Matching Mechanism by Tadelis and Zettelmeyer (AER, 2015)

The Anatomy of a Large-Scale Hypertextual Web Search Engine by Brin and Page (with Appendix A: Advertising and Mixed Motives)

Freakonomics Radio Ep 441: Does Advertising Actually Work? (Part 2: Digital)

High Signal podcast

Watch the podcast episode on YouTube

Delphina's Newsletter

Chris Fonnesbeck, veteran analyst for the Yankees and Mets and creator of the open-source Bayesian modeling library PyMC, joins to unpack why baseball has been a leading indicator for data science and analytics for over a century, and why builders and AI leaders need to pay attention now. The reason it has led is simple: huge incentives and a culture that treats decisions as quantifiable. With wins worth about eight to ten million dollars apiece and front offices built around probabilistic reasoning, baseball has had every reason to push the methods further and faster than industry.

The skillset and culture that built this lead is what AI teams now need to adopt more of: probabilistic thinking, hierarchical models, integrating expert judgment, reasoning carefully under uncertainty, and increasingly causal inference. The conversation traces the throughline from those early statistical innovations to the decisions driving multi-million dollar contracts today, with concrete patterns AI builders can take back to their own work: how to handle small samples and high stakes, why outcomes are the wrong thing to measure, what changes when you push uncertainty all the way through your model, and why robust causal inference needs to be the next frontier.

LINKS

Chris on LinkedIn

The Signal and the Noise: Why So Many Predictions Fail--But Some Don't by Nate Silver

Superforecasting: The Art and Science of Prediction by Tetlock and Gardner

The Book: Playing the Percentages in Baseball by Tango, Lichtman, and Dolpin

High Signal podcast

Watch the podcast episode on YouTube

Delphina's Newsletter

Noah Bruegmann, President of Data CRT, joins High Signal to discuss how to move your data function from a cost center to a strategic "value center". He explains how AI amplifies your existing data culture, the importance of "no-assistance" reporting, and how rebranding documentation as "Context" can finally secure executive buy-in. Drawing on 15 years of experience spanning trading floors and Silicon Valley startups, Noah argues that for too long, data teams have been submerged under an "iceberg" of invisible data preparation. He details how the arrival of LLMs and agentic tools is fundamentally shifting this landscape, automating technical drudgery and allowing data professionals to transition into what he calls "Jack Ryan" mode: acting as high-level intelligence analysts rather than mere number crunchers.

We dig into the architectural and psychological shifts required to navigate this new era and why the most valuable skill in an AI-augmented world is no longer mastering SQL syntax, but "problem framing": the ability to reduce business ambiguity into high-leverage insights. Noah cautions that while AI offers a dopamine hit of instant answers, it demands a new discipline of rigorous verification to avoid automated hallucinations. The conversation provides a clear directive for executives: move past the "ticket-taker" model and start treating the data team as the essential "left-side brain" for organizational decision-making.

LINKS

Noah on LinkedIn

High Signal podcast

Watch the podcast episode on YouTube

Delphina's Newsletter

Jordan Morrow, SVP of Data & AI Transformation at AgileOne and the "Godfather of Data Literacy," joins High Signal to discuss the shift from being "data-driven" to becoming "AI-enabled." Jordan warns that many organizations are walking into the same traps that derailed the data science era a decade ago: prioritizing expensive tooling and hype over the cultural change and literacy required to actually move the needle. The pattern is already visible: enterprise AI projects are failing at around 90%, while individuals doing shadow AI are quietly thriving. The catch is that shadow AI brings its own risks, with people feeding sensitive data into public models without governance. He argues that because AI is probabilistic rather than deterministic, the bottleneck for success has shifted from technical coding ability to a user’s ability to apply "Engineered Intelligence," a blend of machine capability and human emotional intelligence.

LINKS

Jordan's new book "Data and AI Skills: Gain the Confidence You Need to Succeed" (also here on Amazon)

Jordan on LinkedIn

High Signal podcast

Watch the podcast episode on YouTube

Delphina's Newsletter

Dawn Woodard (Distinguished Engineer, LinkedIn), Andrés Bucchi (LATAM Airlines), and Jeremy Hermann (CEO & Co-Founder, Delphina) join High Signal for a deep dive into the shifting architecture of data science & analytics in the era of AI. As the industry moves from static dashboards to vibe coding and conversational querying, this panel of industry veterans explores why traditional data fundamentals—strict cataloging, verifiable outputs, and a single source of truth—are suddenly the most critical bottlenecks in the AI era.

We dig into the sobering reality of the "source of truth" problem, where the speed of AI-generated code far outpaces our ability to define what "correct" actually looks like in a complex enterprise. The conversation reveals how AI is breaking legacy experimentation platforms, the transition of the data analyst into a "verifier" of AI-generated workflows, and why "headless" security architectures are essential for the next generation of autonomous agents. From the limitations of LLMs in causal reasoning to the challenges of integrating AI into "brownfield" enterprise codebases, this discussion provides a grounded framework for leaders navigating the gap between AI hype and operational reality.

LINKS

Dawn on LinkedIn

Andrés on LinkedIn

Jeremy on LinkedIn

Build Your Bracket with Data in the Delphina Sandbox

High Signal podcast

Watch the podcast episode on YouTube

Delphina's Newsletter