Many reasonable people believe that artificial intelligence is about to have a bigger impact on our world than any development since the Industrial Revolution. Societal and economic change at this scale will inevitably bring with it an unlimited number of ethical challenges. If you are interested in AI and its attendant ethical implications, you will not soon run out of brand-new, hot-off-the-presses books to peruse, often containing messages that are diametrically opposed to those of the book you just finished reading. A recent addition to the AI Ethics collection is Carissa Véliz’s Prophecy: Prediction, Power, and the Fight for the Future, from Ancient Oracles to AI (2026). Véliz is an associate professor in Philosophy at the Institute for Ethics in AI at Oxford University.

Véliz’s previous book, Privacy is Power: Why and How You Should Take Back Control of Your Data (2022), paves the way for this one even though she did not address AI in that book. In it, she argued that privacy should be important to all of us because when we cannot control our information, others have power over us. The others we need to worry about, she strongly believes, are the Big Tech companies like Google and Meta that have learned that our information can be very profitable for them. The key message is that largely unregulated huge social media companies profit inordinately by accessing, selling, and otherwise controlling our private information. Worse, that control over information undermines our democracy in many ways, especially by threatening authoritarianism because Big Tech can and often does aid and abet Big Government.

In her new book Prophecy, Véliz’s attention turns from privacy to prophecy (prediction). This book is not as tight as her previous one. It could have been substantially shorter, but much of the unnecessary material it includes is interesting though not integral to Véliz’s thesis. The key message in this book is that similar (and often the same) Big Tech companies (Google, Meta, Microsoft, OpenAI, Anthropic, etc.) that have now dived into AI are threatening humanity’s best interests by promising to tell our future, something humans have always found appealing.

The book’s first two chapters are thought-provoking, though Véliz explicitly tells the reader that they may be skipped. Chapter One covers lots of material–from Roman emperor Tiberius (who apparently had court astrologers thrown off cliffs if their predictions failed) to a 2022 survey indicating that 37% of Americans between the ages of 18 and 29 said they believed in astrology. The second chapter is primarily about the mathematization of predictions over the centuries (by mathematicians such as Bayes, Laplace, and Condorcet), which led to the birth of the insurance industry among other developments.

Chapter Three introduces AI as “The Ultimate Prediction Machine.” Thereafter the book ranges widely, but contains a worthwhile and interesting examination of an assortment of well-known complaints about AI, and especially its predictions:

  • AI models are not reasoning through to a conclusion, they are merely predicting the next word
  • AI models hallucinate
  • AI algorithms’ predictions often reflect the biases of the real world
  • AI models’ predictions can create self-fulfilling prophecies
  • Large language models are fortune tellers more than truth tellers
  • AI creates risks by projecting an illusion of accuracy for its often wildly erroneous predictions

Although most of Véliz’s criticism of AI are familiar, her arguments are well articulated and supported, often with unique reasoning and creative examples. She writes about Nietzche, quantum mechanics, Tolkien, chaos theory, Hannah Arendt, Sam Bankman-Fried, effective altruism, billionaires’ desire to live forever, and much more. As noted, this is a commodious book.

Bottom line: This is a worthwhile volume. Let’s close by quoting two passages from Véliz’s “Postscript”:

“To develop AI ethics is not to give up on regulation—quite the contrary. Good laws are based on ethics. The greater our understanding about AI ethics, the better laws we will have, and the more we will respect those laws. Anyone who has a background in medical ethics recognizes the importance of having both ethics and good laws.” (p. 290)

 

“Prediction is digital technology’s other great sin [in addition to invasion of privacy]. If we weren’t using AI to make predictions about people, AI wouldn’t have as many problems with bias and discrimination, and it wouldn’t have a tendency to fabricate information, thereby being the perfect tool for misinformation. More important, if we weren’t trying to predict people’s lives, we wouldn’t be as interested in surveilling them, and thereby controlling them.

“Surveillance and prediction are partners in crime in the struggle for power. There is no democracy without citizens having agency over their own lives. In being designed to misappropriate our autonomy, surveillance and prediction are digital technology’s original sins. “(p, 291)

Food for thought.

 

Sources

Jennifer Szalai, “Book Review: Fortune Hunters,” New York Times, June 14, 2026.

Carissa Véliz, Prophecy: Prediction, Power, and the Fight for the Future, from Ancient Oracles to AI (2026).

Carissa Véliz, “The Surveillance Delusion,” in The Oxford Handbook of Digital Ethics (Carissa Véliz ed., 2024).

Carissa Véliz, The Ethics of Privacy and Surveillance (2024).

Carissa Véliz, “Losing Skills,” in AI Ethics (David Edmonds, ed. 2024).

Carissa Véliz, Privacy is Power: Why and How You Should Take Back Control of Your Data (2022).

Louise Wilsdon, “Book Review: Carissa Véliz, Privacy is Power, International Data Privacy Law (Vol. 12, No. 3, Aug. 2022).

Videos

“Running with Scissors: AI and the Race for the Future,” at https://ethicsunwrapped.utexas.edu/video/running-with-scissors-ai-and-the-race-for-the-future.