Baggini reviews Carissa Véliz's new book Prophecy: How Anticipating the Future Shapes the Present, arguing that the real hazard of forecasting systems is not that they are wrong but that they become self-fulfilling.
Véliz's thesis: the more a society trusts a prediction, the more it reorganizes itself around that prediction, until the prediction becomes the only outcome the system can produce. Credit scores are the everyday example; AI-driven medical triage is the emerging one.
Baggini notes the book's sharpest chapter is on oncology — where predictive models “determine who will receive treatment and who won't,” and where a single biased training set can quietly eliminate an entire subpopulation from the pool of candidates for experimental therapies.
Why It Matters For Portfolios
The healthcare innovation sleeve we lifted to 4% last week sits directly in this territory. UNH beat on Q1 today (stock up 5%), but the longer story in insurance and provider networks is algorithmic triage — who gets the PET scan, who gets the immunotherapy, who gets hospice.
Véliz's warning is not anti-technology; it is anti-passivity. The firms that will win this cycle are the ones that build auditability and recourse into the model, not the ones that optimize for throughput.
What This Means For Our Portfolios
- Healthcare Innovation 4% — stays. UNH, LLY, VRTX, ISRG. Post-earnings, UNH is the bellwether.
- AI/Cyber 20% — watch the audit-trail names. PANW, CRWD, NET pricing in the next regulatory cycle; MSFT Azure's governance tooling is the quiet winner.
- Avoid pure-play prediction shops. If the regulatory pendulum swings per Véliz's argument, the most exposed names are the ones that sell “risk scores” without recourse.