Power and Prediction: The Disruptive Economics of Artificial Intelligence

Summary of Power and Prediction

1. 1

   When Prediction Becomes Cheap, Everything Changes

   

   00:00

   00:00

   In 2013, Monsanto paid $1.1 billion for a company most people had never heard of. The Climate Corporation wasn't developing miracle seeds or new pesticides-it was replacing something farmers had relied on for generations: their own judgment. Using AI to analyze weather patterns, soil conditions, and crop data, the company could tell farmers exactly when to plant, when to harvest, and what would happen if they didn't. This wasn't just a helpful tool. It was a fundamental shift in who held the knowledge-and the power. This is the heart of what's happening with AI today. We're living in what might be called "The Between Times"-that awkward period after a technology proves it works but before it actually changes how we live. Think about electricity: Edison's light bulb debuted in 1879, yet only 3% of American homes had electricity twenty years later. It took another two decades to reach 50%. Those forty years weren't about perfecting the technology-they were about rebuilding entire systems around it. AI is following the same pattern, and understanding why reveals something profound about how transformation actually happens.

2. 2

   The Real Cost of Knowing What Comes Next

   Start Listening

   AI's breakthrough isn't thinking - it's making prediction radically cheaper. In 2012, a University of Toronto team revolutionized image recognition by treating it as a prediction problem: What would a human say is in this picture? Their AI eventually surpassed human accuracy. We predict constantly: checking weather before dressing, estimating travel time, evaluating job candidates. Better predictions enable better choices. But predictions aren't enough. Judgment - determining what outcomes we actually value - remains entirely human. In *I, Robot*, a robot saves Detective Spooner instead of a drowning girl, calculating his 45% survival chance versus her 11%. The robot valued all lives equally. Spooner believed the girl's life was worth more - a judgment call no algorithm made. When deploying AI, we must be explicit about our values, because machines optimize for whatever we specify, without questioning whether we've chosen wisely.

3. 3

   The Hidden Rules Running Your Life

   Start Listening

   Barack Obama wore only gray or blue suits as president. When asked why, he told *Vanity Fair*: "I don't want to make decisions about what I'm eating or wearing. Because I have too many other decisions to make." This wasn't quirky preference-it was strategic cognitive conservation. Most of our lives run on autopilot through rules, routines, and habits that let us ignore information entirely. Nobel Prize-winning economist Herbert Simon called this "satisficing"-accepting good enough rather than optimal. We adopt rules not because they're perfect but because constantly processing information is exhausting. Two factors determine whether we actively decide or default to rules: the decision's consequences and the cost of information. When stakes are low or information is expensive, we default to rules. AI fundamentally disrupts this balance by making information cheap. During the 2014 London tube strike, 5% of commuters permanently changed routes after being forced to experiment-discovering alternatives that saved six minutes daily. They'd followed suboptimal rules for years simply because finding better information was too costly. When information becomes accessible, we can replace rigid rules with informed decisions.

4. Chapter 4

   The Uncertainty We've Learned to Accept

   Economist George Stigler once quipped, "If you never miss the plane, you're spending too much time in airports." Yet modern airports have transformed into destinations with spas, casinos, and high-end shopping - people now spend an hour longer at airports than a decade ago because travel uncertainty has increased. The wealthy who fly privately experience an entirely different reality. Private terminals are spartan because there's no waiting - the plane leaves when passengers arrive. Without uncertainty, there's no need to make waiting pleasant. This reveals something crucial: rules arise because embracing uncertainty is costly, but they create their own problems. AI could transform airports by addressing traffic and security uncertainties. Navigation apps already estimate travel times; future versions could account for actual flight departures and predict security line wait times, letting travelers make informed decisions rather than relying on conservative rules that waste time. But here's the catch: businesses profiting from waiting passengers have little incentive to eliminate wait times. The Shirky Principle states that "institutions will try to preserve the problem to which they are the solution." Finding AI opportunities requires looking beyond the guardrails protecting rules from uncertainty - and recognizing who benefits from maintaining those rules.

   Chapter 5

   When the Glue Holding Systems Together Dissolves

   Rules are the glue holding complex systems together. Atul Gawande's *The Checklist Manifesto* shows that even skilled specialists need checklists - Boeing's Model 299 bomber succeeded only after implementing one. These procedures ensure reliability, but each represents hidden uncertainty that AI prediction could address. The problem is that rules treat everyone identically despite fundamental differences. Radio stations broadcast the same songs to all listeners; streaming services create personalized playlists. Pandora researchers discovered they could use AI to predict how much different users disliked ads, enabling personalized ad frequency rather than uniform rules. Education is filled with uniformity-creating rules - from seating arrangements to homework policies. Economists Jin and Sun used AI to personalize entrepreneurship training for e-commerce sellers, analyzing each seller's operations and recommending specific modules. Revenue increased 6.6% - transformation, not just efficiency. Long-standing rules become so embedded that everything must move simultaneously for change to occur. Television advertising rules dictate program length and commercial breaks. YouTube's AI-driven system allows content of any length and matches viewers with content and ads they'll enjoy - far more valuable than in rigid network television. The difference isn't the technology - it's the system built around it.

   Chapter 6

   The Power Question Nobody Wants to Ask

   When David Friedberg launched The Climate Corporation, his AI relocated decision-making from rural farmers to San Francisco technologists. The system showed precise field conditions and optimal planting times, replacing farmer expertise and autonomy. As Michael Lewis observed, no one asks: "If my knowledge is no longer useful, who needs me?" This is AI's real disruption. When Blockbuster recognized they needed a Netflix-like subscription model, franchisees earning 40% of revenue from late fees resisted successfully. The company died from internal power struggles, not ignorance. AI doesn't transfer decisions to machines-it shifts which humans make decisions, dramatically altering power dynamics. This explains why system-level AI adoption is so difficult. The COVID-19 pandemic illustrated this vividly: by January 2021, only 9 million Americans had COVID-19, while 320 million faced restrictions. AI-powered cough detection and rapid tests could reliably identify infectious individuals, yet implementing workplace testing required removing rule-based barriers and committing executive resources to system change. The technology existed; the system didn't.

   Chapter 7

   The Future That's Already Here, Just Unevenly Distributed

   AlphaFold, an AI predicting protein structures, has been called "the most important achievement in AI - ever." But AI's greatest potential lies in transforming innovation itself. Innovation involves structured trial and error: specify objectives, generate hypotheses, design experiments, learn from failures, deploy at scale. AI generates thousands of possible solutions from existing data - enabling innovation with higher yield and greater ROI. Team New Zealand's America's Cup victory demonstrates this. Partnering with McKinsey, they developed AI sailors running hundreds of simulations in the time humans ran a handful. After eight weeks, these AI sailors outperformed humans and taught them new techniques. This is the system mindset - recognizing that generating real value requires reconstituting entire systems involving both machine prediction and humans. Leading tech companies like Amazon and Google haven't simply replaced tasks with AI - they've built completely new systems. The question isn't whether AI will transform your industry. It's whether you'll be the one doing the transforming or the one being transformed. In The Between Times, the greatest risk isn't moving too fast - it's moving too slowly while someone else rebuilds the system around you.

   Continue Reading