Welcome to Revise and Resubmit, where we take complex ideas and make them accessible, one book at a time. This is our Weekend Book Review, and today, we're diving into Differential Privacy by Simson L. Garfinkel—published by The MIT Press on March 25, 2025.
Simson Garfinkel is no stranger to the world of data privacy. A researcher at the crossroads of AI, privacy, and digital forensics, he’s a fellow of the AAAS, ACM, and IEEE. His work has shaped the way we think about protecting personal information in an era where data is currency. And in this book, he takes on one of the most debated topics in modern data security—differential privacy.
So, what is differential privacy? In simple terms, it's a mathematical approach to protecting personal data by adding just the right amount of statistical noise. Google, Apple, Microsoft, even the U.S. Census Bureau have all adopted it. But here’s the catch—while differential privacy offers strong privacy guarantees, it also sparks heated debates. How much noise is too much? Can we truly balance data utility with privacy? And are we ready to trust algorithms with decisions that affect millions of people?
Garfinkel walks us through the history, the controversies, and the real-world applications of this technology. He explores why the Census Bureau chose it for the 2020 count and why some critics still argue that it may not be the silver bullet we hope for.
As we unpack this book, one question lingers—are we safeguarding our privacy, or just making it harder to see the truth?
A big thank you to Simson Garfinkel for this insightful work. If you enjoyed this episode, don’t forget to subscribe to Revise and Resubmit on Spotify and check out our YouTube channel, Weekend Researcher. You can also find us on Amazon Prime Music and Apple Podcasts. See you next time!
Reference
Garfinkel, S. L. (2025). Differential Privacy. The MIT Press. https://doi.org/10.7551/mitpress/15354.001.0001
Youtube channel link
https://www.youtube.com/@weekendresearcher
Support us on Patreon
https://patreon.com/weekendresearcher
