Reading
Favorite books that have influenced my thinking.
Pieces of the Action
Vannevar was FDR's science advisor during WWII, and shepherded tech from penicillin to proximity fuses to the atomic bomb. His advice on aligning scientists with the hierarchical war effort applies to any tech org. The challenges he saw the US facing in the 1970s, compared to the progress we have made and what we face today, is heartening. His vision for the future of personal computing (the Memex) was prophetic if incomplete. Written in 1970 and republished in 2022, it could have used a firmer edit, but Bush's voice and anecdotes shine through. I gained an entirely new perspective on innovation and its messy deployment in practice.
Causal Inference in Python: Applying Causal Inference in the Tech Industry
Teasing causality from data is one of the most subjective and challenging tasks that data scientists face. This book provides a thoughtful, fun to read, and practical introduction to causal inference. It covers a lot of ground, from causal graphs to synthetic controls to ML estimates for user level treatment effects. By weaving together theory and code, Facure turns what’s often an abstract topic into something directly usable in industry settings. The math notation is not always clear, but the code examples and reasoning are fantastic. Much of what's here I had to figure out over the course of my career. Other ideas were entirely new to me - and I can't wait to apply them.
Abundance: What It Takes to Build
I found this book to be timely and exciting, as it is written by respected journalists from the left and clearly diagnoses the failures to support the supply side required for achieving Abundance as a society. They convincingly argue that homelessness stems from a simple fact: we don’t build enough housing. They identify many of the factors crippling our ability to build in liberal cities. They illustrate how well-meaning special interests on the left can cripple legislation like the CHIPS act. They explain the history of legal tools created to fight abuse of the commons in the 60s and then explain how those same tools are now used by all to obstruct progress. They critique our federal science apparatus for being too risk averse and bureaucratic. I didn't leave the book with a clear blueprint for what should come next, but I hope this book inspires the start of one.