Before the Story Settles

What Is Go — and Why Did It Matter So Much?

Go originated in ancient China more than 2,500 years ago. It is played on a 19×19 grid with black and white stones. The rules fit on a single page. The number of possible board positions exceeds the number of atoms in the observable universe — roughly 10¹⁷⁰ compared to 10⁸⁰. Chess grandmasters can be beaten by brute-force calculation. Go cannot. It requires intuition, pattern recognition across the whole board simultaneously, and something that looks, to human players, indistinguishable from creativity.

For this reason, AI researchers had long held that Go was the final frontier — the game a machine could not master without something approaching human understanding. The consensus in 2015 was that it would take at least another decade.

It took one year. AlphaGo, built by Google DeepMind, defeated Lee Sedol — the greatest Go player of the previous decade — 4 games to 1 in Seoul in March 2016. The match was broadcast live across South Korea, China, and Japan to audiences of hundreds of millions.

Move 37 — The Moment Nobody Expected

In Game 2 against Lee Sedol, AlphaGo played Move 37 — a stone placed in a position so unexpected that commentators initially called it a mistake. No human player in the thousands of years of Go history had made such a move. It defied every established principle of the game.

It was not a mistake. AlphaGo won the game. Move 37 has since transformed how professional players understand Go. It came from a machine that had taught itself to play by playing millions of games against itself — a form of recursive self-improvement that is the direct ancestor of the agential AI now entering every knowledge workplace.

China's Sputnik Moment

AI researcher Kai-Fu Lee — former head of Google China — called AlphaGo's victory China's "Sputnik moment." Just as the Soviet satellite launch in 1957 shocked the US into an arms race for space, AlphaGo's defeat of first Lee Sedol and then Chinese world champion Ke Jie in May 2017 shocked China into a national AI mobilization. Within months, the Chinese government released its New Generation AI Development Plan — committing to global AI leadership by 2030.

What AlphaGo Actually Was

AlphaGo was not programmed with Go strategy. It was trained using reinforcement learning — playing games against itself, millions of times, improving with each iteration. Its successor AlphaGo Zero started with no human knowledge at all and surpassed AlphaGo within 40 days. This is the recursive self-improvement the fourth question in our survey asks about. AlphaGo demonstrated it was real, it was fast, and it produced capabilities no human had imagined.

The Race That Followed

The US-China AI arms race that is now justifying $725 billion in annual infrastructure spending — the data centers scraping deserts and draining aquifers documented in these books — traces directly to that Go board in Seoul. The ten corporations in the table below all read the same lesson from Move 37: the machine that teaches itself surpasses everything humans thought they knew. The race to build that machine is the Gyre. The communities beside its data centers are the sacrifice zones.

What Ke Jie Did After

Ke Jie wept after his loss. He said he had played what he believed was a perfect game, and AlphaGo had still won. He retired from professional Go in 2023, citing the impossibility of competing in a world where human mastery had been rendered irrelevant. Lee Sedol retired even earlier, in 2019, saying: "Even if I become the number one, there is an entity that cannot be defeated." Two of the greatest human players in a 2,500-year-old game walked away from it because a machine had made their life's work feel smaller. This is the antenarrative the Storying Economy must hold: not just the jobs, but the meaning.

Term One

Artificial Intelligence (AI)

What it is: Software trained on vast datasets to recognize patterns, generate text, images, code, or decisions — without being explicitly programmed for each task. The AI in your phone's autocomplete, your email spam filter, a medical imaging scan reader, and the language model writing this sentence are all AI.

What it is not: Conscious. Self-aware. Capable of wanting anything. Current AI systems, including the most powerful language models, have no goals of their own. They optimize for the objective their designers specified — which is exactly why the designers' values matter so much.

"AI" in 2026 is doing everything from diagnosing cancer to targeting artillery. The word covers a vast range. When a corporation says "AI will create jobs," ask which AI, doing what, for whom, at whose cost.

Term Two

Artificial General Intelligence (AGI)

What it is: A hypothetical AI system that can perform any intellectual task a human can — and transfer learning between domains the way humans do. An AGI could, in principle, teach itself to be a lawyer in the morning and a structural engineer in the afternoon, without being retrained for each. No AGI exists yet. Every major AI lab claims to be building toward it.

Why it matters now: The $725 billion being spent on AI infrastructure in 2026 is explicitly justified by the race to AGI. The sacrifice zone logic — scraping deserts, draining aquifers, displacing communities — is rationalized as necessary for a technology that does not yet exist and whose timeline no one can honestly specify.

The AGI race is the arms race the books document. The Tourbillon and the Drizzle tracks the bubble built around it. The communities stopping data centers are, among other things, stopping infrastructure being built for a technology whose benefits — if they arrive — will flow entirely to those who own the infrastructure.

Term Three

Agential AI (AI Agents)

What it is: AI systems that don't just answer questions — they take actions. An AI agent can browse the web, execute code, send emails, manage calendars, make purchases, file documents, and coordinate with other agents — all without a human approving each step. In 2026, every major corporation in the table above is deploying AI agents to replace knowledge workers at scale.

What makes it different from AI: Agency. A language model answers. An agent acts. The gap between those two things is the gap between a calculator and an employee — except the agent never sleeps, never unionizes, never asks Mahavira's question, and is owned by the corporation that deployed it.

When Oracle says it is "replacing humans with AI agents," this is what it means. Not a tool that helps a worker — a system that replaces the worker entirely. The layoff numbers in the table above are, in most cases, agential AI displacement. The number not yet counted is much larger.

The Method of This Book Series

Antenarrative Phenomenology

Antenarrative — from the Latin ante, meaning before, not against. An antenarrative is the living story that exists before the official version is written, before the consultant summarizes it, before the IRB seals it. It is fragmentary, non-linear, unresolved — the texture of experience before the institution decides what it means. The term was originated by David Michael Boje beginning in the 1990s and is now a cornerstone of organizational storytelling theory. Full theoretical grounding: antenarrative.com

Phenomenology — the philosophical tradition, originating with Edmund Husserl, of attending to experience exactly as it is lived — before categorization, before summary, before the official story forecloses it. Husserl called it "returning to the things themselves." Heidegger extended it: we are always already thrown into a world we did not choose. Harold Garfinkel brought it to ordinary social life — watching how communities construct meaning in real time, before any theorist arrives to explain it to them.

Antenarrative phenomenology, as practiced in these books, is the recording of living experience — from a jog beside a draining lake, from a Socorro community meeting, from a county commission chamber — before the outcome is sealed. Arihanta studied Husserl with Louis Pondy at Illinois, worked with Harold Garfinkel at UCLA, moved through Heidegger at LMU and NMSU, and arrived at this synthesis across decades of morning practice. The jog is the method. The lake is the data. The community voices are the archive.

The Scientific Warnings — Two Open Letters

What the People Who Built It Are Saying