AI-Proof: What Machines Can Never Be — Volume I: Awakening, Part III: The First Awakening

Where we are: after Part II, the start of Part III, the last Part of Volume I. From the machine to the practice of working beside it.

← Previous: Chapter 8: Understanding the Machine You’re Working With · Series hub · Next: Chapter 9: The Architect Mindset →

Priya did her best work on a module already being thrown away. A machine makes you faster at anything, including the wrong thing. Part III is about working beside it: framing, depth, judgement, checking, and change.

In Brief: For readers who want to work beside the machine and keep the work their own. Part III turns from the machine to the practice of working beside it. Its five chapters ask what the work is for and who decides, what depth is when anyone can have the knowledge, and when judgement should yield. They ask how to use the machine and stay able to check it, and what carries a person through change. Every practice names who must give its conditions, and none promises a wage.

Each practice has a person’s half and an institution’s half.

Opening: The Wrong Module

Priya Chandran spent two weeks on an excellent design for a part of her company’s product. The company had already decided to retire it. A colleague told her in a corridor, apologetically, assuming she knew. She did not know.

A machine would have made her faster, and she would only have reached the wrong place sooner. Part III begins there. It is about the work a person does beside the machine, once she knows what the machine is.

The question this part asks

Once you know what the machine is, how do you work beside it so that the work stays yours, and who must give you the conditions to do it?

Part I asked what is yours, and Part II what the machine is. Part III is about practice, and about the limits of practice. Each chapter offers one exercise, and each says what its exercise cannot do.

The chapters

Chapter 9, “The Architect Mindset”, separates three kinds of attention. The operator runs the work, the governor watches it, and the architect asks what it is for. None of them is a rank, and good work often needs all three. The ability to frame a problem does not, by itself, give the authority to decide which problem should be solved. Aristotle, Goodhart’s law, Donella Meadows and the four questions Indian scholars asked before any treatise all help. So do Rajan, his floor supervisor in Coimbatore, and a brand strategist in São Paulo. Priya now writes one question on a card every Monday. By the chapter’s end it has led her to send one message, and nobody has yet replied.

Chapter 10, “Building Deep Human Capital”, opens at the back of a classroom in Sagar, where Kavitha Sharma watches a platform teach her class. Depth, it argues, is what knowledge does to the knower when it is heard, tested and lived with: śravaṇa, manana and nididhyāsana, Yājñavalkya’s three steps. The machine is the best hearing there has been. It can help or harm the testing, and it cannot do the living with for you. A toolmaker in Nagoya, a planner in Medellín and a sensor engineer in Shenzhen find their depth tested. Depth fills a position; it does not grant one. A girl’s question about belief and seeing is still unanswered at the end.

Chapter 11, “Judgement: The Skill That Compounds”, follows Arjun Mehta, whose beautiful website did not work for his client’s guests. An editor in Lisbon, a manufacturing director in Pune and a logistics owner in Lagos whose best decision was luck join him. Judgement is not the override. It grows from mistakes judged by what was known, not by how things turned out. Where decisions are bounded, measurable and recoverable, the machine should often decide, and someone answers for letting it. Four things stay apart: the quality of a decision, judgement as a capacity, the authority to decide, and the duty to answer.

Chapter 12, “Becoming AI-Leveraged”, is the daily practice. Priya checks every recommendation in a report and none of the numbers under them, and one is wrong in front of eight people. Four others, from an engineer in Mexico City to a journalist in Accra, show the machine as partner, producer, checker and researcher. A trust ladder of five rungs, from private exploration to acts that cannot be undone, sets how much to check. The slow surrender, seen in a young lawyer in Singapore, is the loss of practice at framing, which does not show in the work. How much of this a person can do depends on who owns the hours the machine saves.

Chapter 13, “The Meta-Skills That Compound Forever”, asks what carries a person through change they did not choose. Kavitha is ordered into a sixty-hour certification. A marine biologist in Da Nang loses her funding to a free platform that serves the fishing communities better. A floor supervisor in Tiruppur reads the grief of two inspectors moved to packing, and pays for the reading. Four dispositions are built by doing and lost by not doing, and no rubric can certify them. Adapting does not mean accepting a goal you have not examined, and sometimes leaving is right. Those who choose a change and take most of its gains should carry most of its cost.

What joins them

Each practice has a person’s half and an institution’s half. Anyone may ask what the work is for, but the seat where purpose is decided is scarce. Depth fills a position, while the time to look and protection for a good refusal are given or withheld. Yielding to a machine is delegation only when someone decided it and can take it back. Checking needs hours, and the saved hour belongs to whoever owns it. The Part never asks a person to question the purpose of their work while others keep the decisions that set it.

Three chapter titles here claim more than their chapters now hold: “The Skill That Compounds”, “Becoming AI-Leveraged” and “Compound Forever”. Each is kept, and each chapter argues with its own title. What compounds is conditional, working well with a machine is a habit rather than a status, and nothing compounds forever.

How this edition reads the title

The print edition subtitled this Part “Becoming the Person the Machine Cannot Replace” in its contents, and “Becoming the Conscious Human in an Unconscious Age” on the Part’s own page. This edition cannot keep either promise. No practice makes a person impossible to replace. And consciousness, which every person has, is not something a practice confers.

What this Part calls a first awakening is smaller. It is the first time a person sees, in their own work, that a decision was theirs and they let it pass. Priya saw it in a corridor, and Arjun in his rejected website. It is the start of a practice, and nobody in Part III arrives.

The uncomfortable question

Every practice in Part III needs conditions a person does not set alone. They include time to check, the authority to refuse, a first rung to learn on, and safety in asking what the work is for. The chapters name who should give them. They cannot make anyone give them.

So a reader whose employer takes the saved hours and gives none back may practise everything here and be paid nothing more for it. The practice may even make that reader more useful to the change that is replacing them. Part III holds that each practice is worth having for what it does, and promises no wage. Whether that is enough, for a person whose hours are not their own, the book leaves open.

Ending: The Neem Tree at the Window

Volume I ends in a staffroom in Sagar. Kavitha has finished the certification a week before the deadline. Her lesson plan scored seventy-one out of a hundred, enough to pass, and her note about the rubric has had no reply. She does not feel triumphant. Under her contempt she has found a fear of being found ordinary, and naming it has made it a little smaller.

Outside, the neem tree stands in the courtyard. She looks at it for a while, and goes home.

Words to keep

• architect: in this book, a person who asks what the work is for, not only how to do it; a way of working, not a rank.

• disposition: a settled state of character, built by repeated acts and lost the same way (Aristotle’s hexis).

• governor: a person who watches a process already run by a machine and knows when to question it.

• operator: a person who does the work as defined, or starts a machine’s process, without asking whether it should be done.

• trust ladder: matching the checking of a machine’s output to what will happen if it is wrong, from private exploration to acts that cannot be undone.

• śravaṇa, manana, nididhyāsana: hearing a teaching; testing it against reasons and doubts; living with it until it becomes the way you see.

Named people in these chapters are composites, and their conversations and circumstances are constructed; the studies, laws and figures cited are real and sourced.

Revised from the print edition (2026)

• What the print edition promised: Part III was subtitled “Becoming the Person the Machine Cannot Replace” in the contents and “Becoming the Conscious Human in an Unconscious Age” on its own page. Its chapters said that the architect’s output “cannot be replicated”, that mastery takes “Ten Thousand Hours That AI Cannot Shortcut”, and that four meta-skills “compound forever”.

• What I now hold: no practice makes a person impossible to replace, and none promises a wage. The five practices keep a position full: framing, depth, judgement that knows when to yield, checking in proportion to what carries your name, and dispositions kept by use. Each needs conditions an institution gives or withholds, and each chapter names who should give them. The title is kept, read as the start of a practice and not a change of state. This overview is new in the online edition.

Next in the series: Chapter 9: The Architect Mindset.