Legal Talk Network

The episode delivers substantial ideas about privacy, power, prediction, and surveillance with reasonable depth. Carissa Véliz articulates several non-obvious frameworks (e.g., privacy as power through the knowledge-power relationship, predictions as unfalsifiable vs. facts, fiduciary duties applied to data brokers). However, the conversation sometimes retreats into explanation of existing concepts rather than generating new insights, and Zach's follow-ups occasionally rehash ground already covered rather than pushing into unexplored territory.

“The more somebody knows about you, the more you are vulnerable to them.”

“predictions are unverifiable. Unfalsifiable. And we are going against the very important tenant of due process in democracy. When we make decisions on the basis of predictions”

Véliz brings a genuinely fresh angle by framing privacy through fiduciary duty law and by separating surveillance from prediction as distinct but linked mechanisms of control. The Oedipus prophecy metaphor and the historical treatment of statistics as 'second best' offer some originality. However, the core 'privacy is power' thesis is her published book's main argument (not novel to the conversation), and some analysis leans on familiar critiques of Big Tech data practices without substantial new frameworks.

“these systems were designed to surveil and predict, which are two fundamental methods of social, uh, control.”

“A prediction can only become a self fulfilling prophecy if we buy it. If we don't buy it, it has no effect.”

Carissa Véliz is an Associate Professor at the Institute for Ethics in AI at Oxford and author of two relevant books on the topic. She has genuine intellectual credentials and domain authority in AI ethics and privacy. However, she is primarily an academic and author rather than a practitioner who has built or operated a company at scale, which limits her to a strong-but-not-exceptional caliber for a B2B audience focused on operators.

“I'm an associate professor at the Institute for Ethics in AI at the University of Oxford. And I'm the author of Privacy is Power and Prophecy.”

“I'm part of the board of the Proton foundation, which is the majority stockholder of Proton, the company”

The episode lacks concrete data, named examples, and measurable specifics. Véliz references Equifax, East Germany surveillance, the Nazis, and Meta rebranding, but rarely provides numbers, timelines, dollar figures, or detailed case studies. The discussion of how battery drain patterns identify users and inferences from music preferences are conceptually interesting but lack empirical backing or quantification. Most claims remain at the theoretical or illustrative level.

“only you in the world has the kind of phone that you have and the exact apps that you have and uses those apps in the exact way that you do”

“between a third and a fourth of the population were being surveilled”

Zach asks reasonable open-ended questions and demonstrates genuine curiosity (e.g., asking about the hope in the book, pushing on fiduciary duties). However, follow-ups are often soft and exploratory rather than challenging or pressure-testing. Zach rarely pushes back on claims, ask for evidence, or probe deeper into tensions. The conversation flows conversationally but misses opportunities for sharper interrogation that would yield more distinctive insights.

“And so some ways it's the power to influence us, um, directly, which is the power to convince us to do something”

“I want to talk about how predictions aren't the power that is prediction”