High Signal: Data Science | Career | AI

Daragh Sibley, Chief Algorithms Officer at Literati and former Director of Data Science at Stitch Fix, joins High Signal to unpack how machine-learning moves from slide-deck promise to bottom-line impact. He walks through his shift from academic research on how kids learn to read to owning inventory and personalization algorithms that decide which five books land in every child’s box. We dig into the moment a data leader stops advising and starts owning P&L-critical calls, why some problems deserve simple analytics while others need high-dimensional models, and how to design workflows where human judgment and algorithmic predictions share accountability. Along the way we talk incentive design, balancing exploration and exploitation in inventory, and measuring success in dollars—not dashboards. LINKS Daragh on LinkedIn Eric Colson on Why 90% of Data Science Fails—And How to Fix It Sudarshan Seshadri on High-Stakes AI Systems and the Cost of Getting It Wrong Delphina's Newsletter

Lis Costa, Chief of Innovation and Partnerships at the Behavioural Insights Team, joins High Signal to explore how behavioral science is reshaping public policy, digital platforms, and machine learning. She explains how defaults influence behavior at scale, why personalization and chatbots are unlocking new kinds of interventions, and what happens when AI systems meet real-world complexity. We also discuss the limits of nudging, the promise of boosting, and why building for human decision-making requires more than just good models. LINKS The Behavioral Insights Team Lis Costa on LinkedIn High Signal podcast Delphina's Newsletter

Sudarshan Seshadri, VP of AI, Data Science, and Foundations Engineering at Alto Pharmacy, dives into the complexities of high-stakes AI in healthcare. He discusses the balance of speed, safety, and uncertainty that influences AI design and its support for pharmacists. The conversation highlights the evolution of data in decision-making, the rise of full-stack pharmacy practitioners using AI, and the essential role of data leaders as architects of irreversible decisions in ensuring patient care and operational efficiency.

Roberto Medri, VP of Data Science at Instagram, discusses the pitfalls of misaligned incentives in product development. He candidly reveals why most experiments fail and how ego can hinder learning. Medri shares insights on transforming Instagram Reels from a struggling feature to a global hit, emphasizing the balance between data-driven strategies and bold leadership. He highlights the importance of experimentation and rapid learning, and explains how embracing humility can drive meaningful impact in tech.

Fei-Fei Li, co-director of Stanford’s Human-Centered AI Institute, dives into the evolution of artificial intelligence, reflecting on her personal journey and the pivotal role of curiosity in shaping innovation. She champions the need for human-centered design in AI, discussing its transformative potential in healthcare, education, and sustainability. Li also explores the concept of spatial intelligence and urges a collaborative approach to build a sustainable AI ecosystem, emphasizing the importance of dignity and community empowerment in technology.

Eoin Mahony, data science partner at Lightspeed and former Uber science lead, shares his insights on effective metrics. He argues that metrics can mislead if their underlying mechanisms aren’t understood, a lesson he learned while optimizing NYC's Citi Bikes. Eoin discusses the pitfalls of relying too heavily on simplistic data and emphasizes the need for rigorous analysis in data-driven decision-making. He also dives into how generative AI can improve workflows while navigating the hype surrounding tech adoption, blending practical advice with real-world examples.

Barr Moses, co-founder and CEO of Monte Carlo, shares insights on the AI readiness crisis many companies face. She reveals high-stakes data disasters, including a shocking $100M schema change. The discussion emphasizes the necessity of data quality and observability, as organizations struggle to align their ambitions with reality. Barr also highlights the transformative role of LLM agents in improving data debugging. Overall, it's a sharp critique of the disconnect between current data practices and the demands of AI.

Tim O’Reilly, founder of O'Reilly Media and a tech thought leader, discusses the dawn of a new era in programming. He argues that AI is enhancing, not eliminating, programming roles, making the field more accessible. The conversation dives into the history of computing revolutions, the impact of venture capital on ride-hailing innovations, and the critical role of community in a decentralized tech future. O’Reilly emphasizes education's shift toward AI skills and advocates for collaboration and ethical practices as we navigate this transformative landscape.

Stefan Wager, a Stanford professor and expert in causal machine learning, dives into the misalignments between prediction and decision-making. He argues that traditional machine learning often neglects the crucial 'what-if' questions businesses face. Stefan shares insights on causal relationships and emphasizes the need for robust experimentation to make informed decisions. He explores the role of causal ML in enhancing customer engagement and optimizing revenue, while also discussing common pitfalls in experimental design, making a compelling case for collaborative learning in the data science field.

Peter Wang, Chief AI Officer at Anaconda and a pivotal force in the open-source data science community, dives into the transformative role of AI in software development. He questions whether engineers will transition from coding to orchestrating intelligence as AI becomes more prevalent. Peter discusses the challenges within the open-source realm and emphasizes the importance of collaboration and adaptability in navigating AI's evolution. He also highlights the need for effective communication between tech builders and business leaders to drive innovation in this new landscape.