A concept born in Cape Town, 1982
A twelve-year-old sits across from an opponent in a school chess tournament in Cape Town.
The position becomes complicated. And somewhere in the middle game, something unusual happens.
He doesn't see the next move. He sees the shape of several possible futures at once
— branching, weighted, alive — and he moves among them rather than down any single one.
He wins the tournament. And he never quite forgets that way of seeing.
For decades the concept lived unnamed — a practice, a habit of mind,
a way of approaching decisions that felt fundamentally different from the
goal-setting and planning everyone else seemed to rely on.
It wasn't until 2010 that he found language for what it was,
and for what was wrong with planning.
He called it Futurizing.
Seven years later, a program called AlphaZero — given only the rules of chess
and nothing else — teaches itself to play in nine hours and defeats
every human-built engine ever made.
When researchers examine how it thinks, they find that it holds multiple futures
simultaneously, weights them honestly, moves among them.
The twelve-year-old had been right all along.
AI had just proved it.
Thainking™
Because AI showed us how.
Futurizing is the ability to hold many possible futures in mind simultaneously and move among them with intelligence and grace. It is not a metaphor. AlphaZero does it literally. Thainking is the practice of learning to do it too.
No account required · Your trees never leave your device · Thainking™ Trademark Application Pending
The Central Distinction
We were taught to plan. Set a goal. Identify objectives. Measure progress. March toward the destination. This works for projects with known variables. It fails in any situation that is genuinely alive — which is most of the ones that matter.
Planning assumes the future is knowable enough to be charted in advance. You set a goal, break it into steps, and execute. Deviation is failure. Uncertainty is the enemy. The further you are from your plan, the more anxious you become.
Planning mistakes the map for the territory. It works beautifully for building a bridge. It collapses in a career, a relationship, a life — anything that lives in genuine complexity and changes without asking permission.
Futurizing does not ask what the destination is. It asks: what are the possible futures from here, how do they feel when weighted honestly, and which direction makes the most sense given what I can see now?
You hold the scenarios in mind at once — some bright, some dark, some unexpected — and you move among them. Not marching toward a fixed point. Dancing through a landscape you can see more of than anyone who is focused only on the path directly ahead.
The External Proof
In December 2017, DeepMind gave a program called AlphaZero nothing but the rules of chess — how the pieces move, and nothing else. No grandmaster knowledge. No opening book. No human strategy. Just the rules. In nine hours it became the strongest chess player that has ever existed.
Every 400 Elo points ≈ 90% expected score in a direct match. The gap between the greatest human player alive and Stockfish — an already superhuman engine built on decades of grandmaster knowledge — was ~518 points. AlphaZero, having never seen a human game, surpassed even that. It did so by beginning with nothing. Not even a heuristic.
Stockfish had been built over years by human programmers and grandmasters. Its evaluation function counted material, rewarded center control, and embodied centuries of accumulated chess wisdom. Even when its opening book was removed for fairness, its core judgment remained shaped by what humans believed made a position good.
AlphaZero had none of this. In several decisive games it deliberately sacrificed pawns and pushed flank pawns for long-term space and piece activity. Stockfish, evaluating with its human-trained system, calculated it was winning. It was not. The strategies AlphaZero discovered — once dismissed as unsound — are now standard in grandmaster play worldwide. The machine had found things the humans had missed for a century.
Silver et al., "A general reinforcement learning algorithm that masters chess, shogi, and Go through self-play," Science, Vol. 362, Issue 6419, 2018The most important result came years later. In a 2025 study published in the Proceedings of the National Academy of Sciences, researchers took AlphaZero's independently discovered concepts — patterns no human had ever named — and presented them as unlabeled puzzles to four of the world's strongest grandmasters, including former and current world champions.
After a single exposure to AlphaZero's patterns, every grandmaster improved significantly on brand-new unseen puzzles. Average gain: +0.85 puzzles solved — statistically significant. One player went from 0/12 to 5/12 correct. The researchers called it "a proof of concept demonstrating the possibility of leveraging knowledge from a highly capable AI system to advance the frontier of human knowledge."
Gao et al., PNAS, 2025 — "Extracting human-interpretable concept-based explanations from deep networks via stochastic perturbation"What AI Revealed
Stockfish was extraordinary — far beyond any human. But its evaluation function was built on human beliefs about what "good chess" looks like. AlphaZero showed that even the ceiling of human-taught intelligence is still a ceiling. The best knowledge we had encoded was still our best knowledge. It could not exceed its own assumptions.
By starting with no assumptions at all — not even a heuristic — AlphaZero explored possibilities that human-taught engines had implicitly ruled out. Not because the moves were secret. Because they conflicted with received wisdom. The pawn sacrifices were always there. We had trained ourselves not to see them.
Daniel Kahneman's decades of work showed us that human thinking is riddled with cognitive biases — confirmation bias, anchoring, loss aversion, availability heuristic. These are not flaws in broken people. They are features of all human minds. The ceiling is structural. Thainking doesn't patch it. It temporarily removes it.
The goal is not to think like a machine. It is to use the machine as a mirror — to see, briefly, what a position or decision looks like without our accumulated assumptions distorting the view. Then bring the human back: the values, the relationships, the ethics, the love for the people around us. The mirror shows what is there. The human decides what to do about it.
The Tradition
For 2,500 years, the greatest philosophical minds worked on the same problem. They lacked one instrument to complete the proof. Now it exists.
Socrates describes prisoners chained in a cave who see only shadows on a wall and take them for reality. One escapes into sunlight, returns to tell the others — they refuse to believe him. The allegory is not about caves. It is about the structural condition of human cognition: we live inside our assumptions and mistake them for the world. Education — and Futurizing — is the climb into light. AlphaZero stepped outside the cave entirely. It had no shadows to mistake.
Tracing every motion backward, Aristotle arrives at a first cause — eternal, unchanging, pure actuality. He calls it "thought thinking itself." The deeper point: all reasoning that does not ground itself in something honest eventually collapses into circular justification. Thainking forces grounding. The tabula rasa reset and the honest scoring of scenarios are the philosophical descendants of Aristotle's demand for a real foundation.
Aquinas builds Aristotle's argument into his Five Ways. His First Way: the chain of cause and effect must terminate somewhere real, or reasoning collapses into infinite regress. For decision-making: every choice that rationalises its way to the answer it wanted from the start is Stockfish — capable, but bounded by what it was already taught to believe. Thainking is the attempt to find a real foundation for each decision, not a convenient one.
Locke: the mind at birth is a blank slate. All knowledge comes from experience — which means we can always start again. Kant corrects him: the mind brings built-in categories (space, time, causality) that structure every experience. Thainking stands at the synthesis: the AI-powered blank slate moment is real and possible (Locke's tabula rasa reset). But we bring our humanity to what we discover in it — our values, our care for others, our ethical judgment (Kant's categories). Neither alone is sufficient.
"Existence precedes essence." We are not born with a fixed nature. We define ourselves through choices. The future is not determined — it is a field of open possibilities. This is the philosophical foundation of Futurizing. Building multiple scenarios and moving among them is not prediction. It is the act of taking seriously that the future is genuinely open, and that how we inhabit that openness is the central creative act of a human life. We are, as Sartre says, "condemned to be free."
Sisyphus pushes a boulder up a hill for eternity. It rolls down. He pushes again. Camus insists: "One must imagine Sisyphus happy." The moment he turns on the descent and chooses to push again with full awareness, he is free. The rock does not change. The hill does not change. What changes is the quality of consciousness brought to it. Thainking operates in exactly this space: you cannot control all outcomes, but you can bring extraordinary clarity to every decision you face.
The Practice
These are not techniques to memorise. They are a repeatable practice that installs Futurizing as your default cognitive mode — the way AlphaZero's training eventually became its nature.
Thirty seconds of deliberate emptying. Before building any scenario, consciously suspend every assumption about what is possible, what is normal, what people like you do. This is the move AlphaZero made by design — beginning with no inherited heuristics at all. Without it, every scenario you build carries invisible weight from your past.
You will feel resistance. That resistance is the sound of the cage door. The reset does not require that you abandon your values. It requires only that you not let received wisdom decide the answer before you have looked honestly at the question.
List four to seven genuine options — including ones that feel unlikely or uncomfortable. For each one, mentally play out two or three believable futures, five to ten years ahead. Score each honestly on how the future actually feels, not how you wish it would feel. Average the scores. Watch the strongest path rise.
This is Futurizing made concrete. You are not planning a route to a destination. You are holding multiple futures simultaneously, weighing them, and letting the picture form. The scoring is not arithmetic — it is a discipline of honesty. The branch you keep avoiding telling yourself is a score of 3.
Socrates walked the streets of Athens asking the question everyone else had stopped asking. He didn't argue — he examined. He turned the question back on the person holding it until the hidden assumption surfaced. The third step asks you to do the same to your own strongest branch.
Not: is this a good idea? But: what would have to be true for this to be wrong? What am I not saying out loud? What future am I quietly refusing to simulate? What would someone who loved me and disagreed with me say? This step is uncomfortable by design. The discomfort is precisely where the hidden flaw lives. Finding it here, in thought, costs nothing. Finding it later, in life, costs everything.
Futurizing sees the whole weather system, not tomorrow's forecast. But most people's scenario trees stop at one or two years — the distance where anxiety lives. The fourth step is the deliberate act of stretching every serious branch to its full length: five years, ten years, the life it eventually becomes.
What kind of person does this path make you into? What does it feel like for the people around you at year seven? What are you trading now, and for what? The Iroquois Confederacy required every council decision to consider its impact on the seventh generation yet unborn — roughly 140 years. You don't need to go that far. But you need to go further than comfort allows. The long horizon rarely changes the available options. It changes the honesty with which you weight them.
AlphaZero distilled millions of games into a small number of reusable concepts — and those concepts transferred to human grandmasters in a single session. You do the same: after every Thainking session, extract one short, named mental model from what you have learned. Write it down. Give it a name.
Over time these concepts become your private library of permanent cognitive upgrades — tools you can draw on in any future decision, and eventually hand to the people you love. What begins as a single decision made more clearly becomes, over years, a living philosophy.
The Engine
The Book
In progress. Fourteen chapters. Purchasers of the 30-Day Protocol receive all chapters as completed.
You're lying awake at 2 a.m. staring at the ceiling. Tomorrow you have to decide whether to take a new job that means moving your family to another country. The pay is better. The schools look stronger. The future feels safer. But your teenager is finally settling in with friends, and your younger child is just starting to love their sports team. What if the move disrupts everything? What if you're wrong about the long-term picture? You run through the usual mental checklist. You've done this a hundred times before. Yet the decision still lands with that same foggy weight. Something important always seems missing.
Begin
The app is free and yours forever. The 30-Day Protocol is the structured education — the practice that turns the tool into a permanent change in how you think.
Payments by Stripe · 7-day refund guarantee on the Protocol, no questions asked · Thainking™ Trademark Application Pending · © 2026 Thainking
API Settings
Get yours at console.anthropic.com
1. workers.cloudflare.com → Create Worker
2. Paste the code below
3. Settings → Environment Variables → add ANTHROPIC_API_KEY
4. Copy the worker URL → paste above
export default {
async fetch(request, env) {
if (request.method === 'OPTIONS') {
return new Response(null, { headers: {
'Access-Control-Allow-Origin': '*',
'Access-Control-Allow-Methods': 'POST, OPTIONS',
'Access-Control-Allow-Headers': 'Content-Type',
}});
}
const body = await request.json();
const res = await fetch(
'https://api.anthropic.com/v1/messages', {
method: 'POST',
headers: {
'Content-Type': 'application/json',
'anthropic-version': '2023-06-01',
'x-api-key': env.ANTHROPIC_API_KEY,
},
body: JSON.stringify(body),
});
const data = await res.json();
return new Response(JSON.stringify(data), {
headers: {
'Content-Type': 'application/json',
'Access-Control-Allow-Origin': '*',
}
});
}
};
Last updated: March 2026
Scenario trees and session history are stored in your browser's localStorage only. They never leave your device unless you purchase a premium plan, which enables encrypted cloud sync.
If you enter a Claude API key, it is stored in your browser only. It is never transmitted to Thainking servers — only to Anthropic's API (or your Cloudflare Worker) when you trigger a consultation. We strongly recommend the Worker approach.
All payments are processed by Stripe. Thainking never stores your payment card details. Stripe's privacy policy governs payment handling.
We do not use third-party tracking analytics. If we introduce any analytics it will be privacy-first and you will be informed.
This site does not use tracking cookies. localStorage is used only to save your own session data locally.
If you purchase the Protocol, your email is stored securely and used only to deliver the programme. We do not sell email addresses. You may unsubscribe at any time.
You have the right to access, correct, or delete any personal data we hold. Contact privacy@thainking.com. EU users have rights under GDPR. Canadian users have rights under PIPEDA.
Thainking is suitable for all ages including children. We do not knowingly collect personal data from users under 13 without parental consent.
privacy@thainking.com
Last updated: March 2026
The name "Thainking," its visual treatment Thainking, and the Futurizing methodology are claimed as intellectual property of Glyn Hotz. Trademark application is pending. Unauthorised commercial use is prohibited.
All philosophical and scientific content is attributed to its original authors and works. No content is reproduced without attribution. All analysis and commentary is original.
Claude AI responses are produced by Anthropic's API. Thainking makes no warranty as to their accuracy. Always apply your own judgment to AI analysis.
Nothing on this site constitutes legal, medical, financial, or psychological advice. All tools and exercises are for personal development purposes only.
If you are not satisfied with the 30-Day Protocol within 7 days of purchase, contact us for a full refund. No questions asked.
AlphaZero and Stockfish are trademarks of their respective owners. This site is not affiliated with Anthropic or DeepMind.
legal@thainking.com