The shape and structure of modern society was not designed — it evolved over thousands of years.
As we invented new technologies, created laws, developed ethical systems, and expanded trade, society has changed, shaped by market competition, wars, empires, and ultimately by the contributions of billions of people.
This complex interaction has resulted in a growing number of ‘layers’ across society — which I think of as the output of a centrifuge — a complex system which separates people into classes over time.
You can think of every company’s talent practices as a small centrifuge, with the interaction of these many small centrifuges, social systems, laws, and international markets combining to produce a relatively consistent global sorting mechanism.
A person being promoted in one company increases their chances at a more senior job in other companies, and the leaders in any particular field tend to cluster — socially, and sometimes, geographically — such as Silicon Valley, for software engineers.
The centrifuge has never been driven by fairness; while individuals and social leaders fight for rights and freedoms, in the long run, the centrifuge responds to market forces, competition, and conflict.
I believe that the faster it spins, the more opportunities people have to both rise and fall. I also believe that demand and competition for top talent will only increase as this process accelerates.
When the centrifuge spins slower than one person’s lifetime, heredity and privilege take precedent over character and ability — the layers ‘settle’ — I believe we’re moving past this.
The centrifuge used to spin every few hundred years, and now it spins multiple times across any individual’s career: the skills people learned in school and university will no longer last their lifetime.
Generative AI will not just accelerate this spin exponentially but also change the characteristics by which people are sorted and reshape society.
This may be AI’s single most impactful use case: if done well, we can provide incredible social mobility, improve fairness, and drive massive economic growth. If done poorly, it could have the opposite impact.
And this is not a zero-sum use case — I’m talking about sustained global economic growth, not simply redistributing existing resources.
When merit is hidden, we don’t incentivise people to perform — we don’t even know who to incentivise. When talent is buried, we fail to apply the most promising minds to the most important subjects.
When our technology gets ahead of our ability to organize society, breakdown occurs. We see unrest, revolutions, war. These crises tilt the centrifuge, forcing the system to sort for different kinds of talent, until the system stabilizes on the next spin.
AI has the potential to create a much fairer and more productive society. Or it could tilt the centrifuge so far that we spin away into nothing…
Who I Am, and Why I Care so Much about Skills & Social Mobility
I was raised by an amazing mother — a single parent, below the poverty line, in a poor town in Australia.
A country known for being lucky.
I now lead Engineering and Innovation for Commonwealth Bank of Australia, where I’ve worked for nearly half my life.
I had two strokes of real luck.
The first was being gifted a computer at age five.
I learned to code because I couldn’t afford software.
Then I learned to hack — because I couldn’t afford the internet.
I had to figure out how to get it for free.
And in doing so, I got to meet some of the best hackers in the world — in the U.S., Russia, Europe, and China.
Some had it better than me. Most had it worse.
And I saw something very clearly:
There’s less correlation between someone’s talent or character and the quality of their life than most people think. We’ve made huge progress towards merit-based equality, but we have a long way to go.
Luck — who your parents are, and where you’re born — still plays a huge role.
That’s why I spend so much of my personal time trying to support free and open technology education, and it’s also why I care so much about what comes next: Because for once, we might have a shot at making the system fairer and more stable, not just faster.
My second big stroke of luck — and this is a weird thing to say — was having cancer. Twice.
Bone cancer at 16, and pancreatic at 32.
When you think you’re at the end of your life, you care a lot less about your title and your bank balance — and a lot more about how you treated people, and whether you’re leaving a better world behind.
The biggest gift, though, was time: months to reflect on what matters, to absorb big ideas, and to observe how people treat each other — what drives them to do their best, and where the fault lines lie. What makes people stronger, and what causes them to break.
Individually and Collectively.
The Balance: How Advances in Technology cause Social Disorder
I think we can measure a global civilization according to its ability to do just three things:
- Its ability to calculate — to measure things and make predictions
- Its ability to create things — to manipulate matter
- Its ability to organize itself — to do things at scale and with cohesion
When these three evolve together, we thrive. When they get out of sync, a bit, society strains. When they diverge, society can collapse.
I believe, when it comes to our ability to organize society, we have always just established enough to survive, as technology evolves and pushes us forward.
And when we fail to organize ourselves well, it’s not just ideas that get lost: it’s people.
The Age of Heredity: Where Leadership Emerges, and Structures Solidify
If you zoom all the way out, the first few spins of the centrifuge, over thousands of years, up until about 1800, gave us a fundamentally two class structure.
In the beginning, society was flat. Our ability to manipulate matter was minimal: sharpened stones. Our ability to calculate was tied to our number of fingers. Our ability to organize ourselves was capped by neurology — Dunbar’s number — 150 people in the village that we could trust, remember, and organize.
And so, society stayed flat. Everyone contributed. Everyone observed. No layers. No hierarchy. Just memory.
Then came a crisis: A threat. A decision. Something the tribe couldn’t face with instinct alone. The centrifuge tilted, and a leader emerged.
Someone who could see further, act faster, inspire others. Get through the crises. Merit made visible.
That leader needed help. Helpers became staff. Roles began to form. Memory was no longer enough — structure appeared.
When the crisis was over, the leaders stayed in place. Then so did their sons. The centrifuge was taking hundreds, maybe thousands of years to spin. Heredity took priority over ability.
Power calcified, but this new structure was superior in many ways. It let us grow beyond 150. It let us discover mathematics, develop language, invent machines, store food. We cut rock and built monuments.
Society separated. Not into many layers: just two. This happened at different times in different geographies, but it was the same pattern globally, between roughly 3000 BC and 1800 AD. Earth converged on a common division: Leaders and Labor.
- On the top: Pharaohs, priests, scribes, nobles — those who claimed divinity, held knowledge, coordinated the labor.
- And below: workers, craftsmen, farmers — some conscripted, some paid, many born into generational service.
Many will point to apparent complexity in these ancient worlds — to Egypt’s court officials, Rome’s equestrian class, China’s scholar-officials, medieval Europe’s intricate feudal hierarchies. But look closer. These were not true middle classes with genuine mobility. They were merely variations within the same binary system — elite power-holders versus laboring masses. Lords might outrank knights, who outranked squires, but the fundamental division remained between those who ruled and those who toiled.
What appeared to be complexity was in fact just elaboration. The merchant who amassed wealth enough to rival a noble was still fundamentally an outsider to power, his position precarious, his children’s status uncertain. The rare peasant who joined the clergy still served the established order. These were exceptions that proved the rule, safety valves that prevented revolution rather than true pathways for mobility.
This was a much more productive form of organization than a single layer model. This rigid two-class system persisted for millennia, the foundation upon which civilizations rose and fell across the globe. By the dawn of the 1800s, the world stood at the precipice of transformation, as whispers of mobility between classes began to emerge.
We pushed the two layer model to its limits: it gave us steam power, complex mathematics, and printed knowledge. But then, the system started to break apart. We’d discovered electricity, and mass production, and the power of an educated workforce — and the old boundaries began to dissolve. The centrifuge was spinning faster now, sorting society with greater precision, creating gradients where once there were only walls.
The centrifuge still tilted from time to time. We had war, we had revolutions. The hype of revolutions exceeded their reality — they promised structural change but only shifted who reached the top layer and what words were used to describe the same things.
The Age of Partial Merit: How the Need for Talent Spun Up the Middle Class
The next few spins of the centrifuge, between the start of the Industrial Revolution, and the bloom of the Internet — created the middle class, and gave us a three-class system, which I believe lasted until about 1990, and many people still think we exist within.
Before about 1800, professions were hereditary. Your father was a blacksmith. You’re a blacksmith. Your last name is your job. My last name is Hopper. I guess my family made beer.
I do not think the emergence of upward mobility was driven primarily by values, or cultural revolutions: these certainly helped, although it is impossible to determine whether this was correlation or causation. I suspect these were an output of the centrifuge, rather than an input.
The driving force powering up the centrifuge? Competitive market forces, and an insatiable need for human talent.
We needed roads. Railways. Ports. Telegraphs. Factories. Restricting the search for talent to the leadership was fatal to many endeavours. Enterprises, companies, and governments had to find more talent — before their competitors could.
We needed the ability to plan. To operate. To design. To track inventory. To coordinate workers. To service machines.
We needed a middle layer, between the rulers and the labor.
For the first time, the system allowed people to move upward. If you could pass the exam, land the apprenticeship, navigate the guild, or build a business — you could rise.
Unions appeared. Merit started to matter — downward mobility was a possibility as well: You had to be good at your job to keep it.
And from this simple combination of market forces, advancing technology, and need for talent, the middle class emerged.
In 1800, just 5% of the world could be called “middle class.” By 1950, it was 20%. By 1990, nearly 45%.
This new mobility wasn’t universal. Many people were given the opportunity to rise, but only those with the right accent, appearance, passport, or network. Women remained largely excluded. Racial hierarchies persisted. Colonial subjects rarely had access to these new pathways. Geography determined destiny as much as ability.
Still, it was better than before. There was some opportunity for people to move up. In around 1990, I’d describe the layers of society very much as most people still think about them now — the upper, middle, and lower classes.
This partial merit system had an enormous impact on economic output and productivity. One generation of elevated talent started training the next — leading to the start of the exponential explosion of technology which we’re bearing witness to today. It let us scale production, services, healthcare, and communication.
In my view, it lasted until the arrival of the Internet, of social media, of mobile phones, of real-time trading platforms. With these technologies, the centrifuge began to spin at speeds never imagined — sorting talent with unprecedented precision while simultaneously blurring the very boundaries it created.
In the last three decades, for the first time, the spin started to be faster than a human career, which means people don’t just have to learn to rise, they need to learn in order to stay in the same spot.
The Information Age would demand something more than partial merit. It would need more than just credentials — it would need persistent performance and career long learning. And most importantly, as we became globally connected, people who were ‘world leading’ — in any field — would take a much larger share of the pie.
The Age of Information: From Three Discrete Classes to Five Blurred Layers
In addition to further accelerating the spin, the Information Age changed the dynamics in a few, very important ways:
- Geography disconnected from top-end services: People started looking for the global best expert or product, not the best locally available.
- Geography disconnected from global renown: You can become famous anywhere, online.
- Specialized communities connected with each other globally: The pioneers in any field were suddenly within reach of each other, leading to them having a compounding advantage against everyone else in their field — if they were good enough to earn their membership.
- Social networks created positive reinforcement loops: Crossing a certain threshold of popularity makes you more likely to cross the next one, not less.
- Early-stage capital separated from the aristocracy: tech venture capital emerged and startups paid in equity, not just cash.
- Global winners emerged: For the first time, companies could ‘win’ at their field globally, not just in their home company or market. Google won search.
- Social systems diverged: Different western economies started treating critical necessities — like healthcare and education — very differently. Some kept them as human rights, and some put them behind a paywall. This tilted the spin differently in different countries.
- The gig economy: Many professions were separated, by technology, from stable employment, and paid by the task instead of by the hour, week, or month.
These made the centrifuge not only spin faster, but have separated working society into five ‘layers’, which merge and flow together, rather than three discrete classes:
1. The Money: Capital aggregators. People who primarily make money by having money — structurally separated from the labor system, but essential to its function. At global scale, capital allocation becomes the only viable mechanism to organize complexity. Not just billionaires — but venture capitalists, sovereign wealth managers, asset allocators, and founders who’ve escaped the need to earn through effort. Anyone who’s crossed the threshold where capital compounds faster than labor, and participation becomes optional. Ring 0.
2. The Pioneers: Adaptive professionals. People who don’t just copy, they make new ideas and push boundaries, participating in the global community of their craft. Always learning, moving, rising. Teaching and sharing.
3. The Credentialed: Execution-oriented white-collar professionals. People who invested a lot of time and effort ‘up front’ in learning a complex profession and then expected those skills to last a lifetime. Essential to running complex systems, but increasingly fragile and resistant to change. Depending on what country they live in, and what profession they picked, they may no longer receive the return on their investment for their tertiary education.
4. The Pillars: The people who shape the physical world and care for others. Builders, mechanics, electricians, nurses, carers, logistics workers. They move matter and sustain life. Without them, none of the other classes would function — they’re the pillars of our societies. They generate enormous social value that markets consistently fail to price. But within their professions, upward mobility still exists — driven by skill, reputation, and trust.
5. The Invisibles: People displaced from stable employment and separated from the protections that it provides, while still doing the same job — e.g., the gig economy. Largely overworked and paid by the task, algorithms, which enforce minimum standards but do not reward excellence, have removed much of their autonomy as well as upward mobility. Very little opportunity for mentorship — any professional growth would have to occur in addition to the mountain of gigs. If the Pillars are holding up society, the Invisibles are the mantle: essential, and with the potential to erupt if the pressure continues to build.
The Scalpel: In Search of the Edge
I’d like to share the story that made me realize how strong the difference is between pioneers and the credentialed. In 2016, I underwent one of the most complex surgeries in modern medicine: a non-pylorus-preserving pancreaticoduodenectomy with partial hepatectomy. The procedure has a high mortality rate — above 5%. My surgeon had performed over 600 and lost just one patient — an elderly woman in her nineties.
I later asked him what made the difference. He said (I’m paraphrasing, but not by much):
“Every year, I speak at conferences about how to do this surgery better. I talk to everyone leading the field. We share advice. I’m never satisfied with how we do it.
Your blood sugar was crashing every fifteen minutes. That made it not just a hard surgery — but a very hard anaesthetic challenge. Tim is the only person I’d work with on that. We’ve worked together for a very long time.”
Pioneers don’t work just for themselves — they work to advance their field.
They learn from adjacent disciplines, respect each other’s craft, and build networks that compound knowledge.
That same mindset — cross-pollinating, experimenting, compounding effort — is what positions them closest to upward movement.
In fact, they’re the most likely group to cross into The Money.
Since 1990, upward social mobility has exploded.
At least seven of the ten richest people in the world today began as Pioneers with more modest means. They leveraged their craft into capital — through timing, scale, and systems that rewarded what they could build.
But the centrifuge works in both directions. For every rise, there’s many potential falls. Entire professional categories in the Credentialed layer have seen their security erode as technology automates their core functions. Travel agents, mid-level managers, certain legal specialists — positions once considered unassailable have vanished or transformed beyond recognition. Likewise, entire segments of ‘The Pillars’ have been disinter-mediated from who they serve.
What’s new isn’t just the stratification — it’s the unpredictability. In previous eras, your starting position reliably predicted your ending position. Today, dramatic movement is commonplace, creating both opportunity and anxiety. We’ve replaced the certainty of stasis with the possibility of rise, but the downside is real.
A major health problem, like the cancer I’ve had twice? In a country without free healthcare, that would have been game over.
Now, let’s talk about what happens when a crisis occurs.
The Tilt: Recalibrating the Centrifuge for Wartime Leaders
When a crisis strikes, the centrifuge tilts. That is, we sort people according to a different formula, looking for a different set of characteristics — more character, less credentials.
We see the Tilt in sport. In companies. During pandemics. In market crashes. If you’ve ever been to a ‘war room’, you know what I’m talking about.
Initially, we keep most people organized the way they were, but we create a separate system for those who we want to separate to handle the crisis response. If the crisis deepens, we shift more and more people under the crisis structure. In extreme cases, like war or catastrophe, the original system is suspended entirely. That’s what martial law is — the final override: everyone must tilt.
And nothing tilts the centrifuge like full-scale war does.
In wartime, we need clarity under pressure, adaptive speed, ability to mobilize others, moral courage, operational decisiveness, and the discipline to act. These take immediate precedence over credentials and charm.
A wartime subclass forms in every layer: defined not by status, but by action and outcomes.
In World War II:
- From the upper class came strategists, statesmen, war financiers
- From the middle class came officers, engineers, analysts, military scientists
- From the lower class came soldiers, fixers, local leaders
And in war, some countries tilt deeper than others. Countries who are attacked and face existential threat tilt further and spin faster than aggressors. More people are willing to mobilize, they mobilize harder, and they mobilize with more autonomy.
Your ability to mobilize depends on the depth of your talent pools, the systems you’ve built to recognize and deploy them, and people’s willingness to engage. Conscripted soldiers rarely outperform volunteers. Forced mobilization lacks the clarity and commitment that the Tilt demands.
Audie Murphy entered WWII as a hungry farm boy, by lying about his age, after being rejected twice. He left it as one of the most decorated US soldiers in history.
That’s the brutal, beautiful clarity of the Tilt: It forces you to see who’s capable, not who’s currently in charge. It forces you to care more about reality, and less about perception.
Short wars are won in preparation. Long wars are won in mobilization. When Ukraine tilted, they mobilized their pioneers — and their pioneers started teaching everyone else.
And this values adjustment, this ‘Tilt’? I suspect that’s why it’s so hard for many people to reintegrate after war. We have psychological techniques to treat trauma. But we have no societal mechanism to help someone re-enter a world that runs on different values and reward systems, particularly when those reward systems seem less fair.
This is what happens when you introduce more dimensions.
A two-dimensional talent assessment only sees confidence and competence.
A four-dimensional talent assessment adds courage and clarity.
When the centrifuge tilts towards war, we stop voting for who is popular, and we start voting for who will keep us alive.
Because in war, status is earned every day. Meaning is clear. Coherence is survival. Merit is visible. The centrifuge spins at full speed.
I believe AI will let us bring this wartime clarity to how we find, monitor, and reward talent — outside of war, and it will move our talent sorting mechanisms from a small number of dimensions into higher-dimensional space.
The Transparent: Visible Versus Hidden Merit
All long-term competition is won by those who can best mobilize human capital. True for sport, true for war, and true for companies. I can’t find a single example of a human endeavour where this is not true. But working out who to mobilize is easier in some professions than others.
Leading cyber security teams, I learned it’s easy to identify great hackers. They break in, show what’s broken. If they can’t, they’re not very good. In cyber-attack: Merit is visible.
But cyber defenders? It’s hard. If nothing goes wrong — is that skill, or luck?
You could have experts and still get breached. Or have amateurs and stay safe.
Merit is hidden.
This divide — between professions with visible merit and those with hidden merit — began to deeply haunt me. I see it everywhere.
And in my experience, professions with visible merit do much better over time than professions with hidden merit: The merit compounds, leading to breakthroughs at an exponential, rather than linear pace.
And AI: Both predictive, and Generative, is going to make talent — and character — more transparent, across almost every industry.
The Potential / Outcome Cycle: Voting, Weighing, and the Loop We’re Trying to Close
Benjamin Graham once said,
In the short run, the market is a voting machine but in the long run, it is a weighing machine.
The same is true inside organizations.
In the short term, we reward confidence, visibility, and narrative.
In the long term, we reward delivery, impact, and truth.
Most modern organizational methods — Agile, OKRs, DevOps, SRE, A/B testing — aren’t just about speed.
They’re about alignment, feedback, and most of all, closing the gap between perception and reality.
These management techniques all try to shorten the cycle between perception — how well we think someone will do — which is heavily influence by attributes like charisma — and actual delivery of outcomes — which is influenced a lot more by followership, real talent, and grit.
Between the vote and the weight.
They spin the small centrifuges — our internal talent loops — faster.
Trying to reward the right things sooner.
Trying to make truth visible before it’s too late.
Trying to elevate the people who will succeed.
And now, AI will push this even further. Organizations will link their talent systems to their product and project systems, and AI will be used to determine who really gets things done. Early correlation is a better recipe than late causation.
This will let us see across the horizon. Not just who is performing now, but who is likely to perform well in the long term.
Not just what went wrong — but what will go wrong.
And maybe, most powerfully of all — who is lifting others quietly in the background.
Consider how this transformation is already beginning: AI can now analyze code contributions not just for quantity, but for elegance, maintainability, and how often others build upon it. It can trace the lineage of ideas through meeting transcripts and collaborative documents, identifying who consistently generates concepts that survive to implementation. It can detect patterns in team communications that predict project success months before delivery dates.
Where traditional management might reward the person who speaks most confidently in meetings (voting), AI can connect those statements with eventual outcomes (weighing). Where current systems struggle to recognize the mentor whose quiet guidance prevents failures, AI can correlate decreased error rates with specific advisory relationships. And where today’s organizations often mistake documentation for progress, AI can differentiate between performative work and substantive contribution.
This closing of the perception-reality gap doesn’t just change how we evaluate existing talent — it fundamentally transforms who we can discover in the first place.
Throughout history, our limited ability to see merit across distance, culture, and convention has meant countless brilliant minds remained unrecognized.
AI doesn’t just tighten the cycle between action and recognition; it expands the universe of talent we can identify.
The Forgotten: Lost, Otherworldly Talent
My favorite example of talent our systems almost failed to see comes from a century before our modern organizational methods: the story of Srinivasa Ramanujan. He was born in 1887 in a small town in India. No formal education, just a mind that could somehow see infinite series and number theory in the way most people see traffic lights and furniture.
He sent a letter filled with equations to G.H. Hardy, a famous mathematician at Cambridge. Hardy almost threw it out, but later said discovering him was “the one romantic incident in my life.”.
Ramanujan wasn’t just good — he was otherworldly. He wrote theorems nobody else had seen, often without proof, but with insight so deep that a hundred years later, we’re still unpacking it.
And we almost missed him.
He wasn’t credentialed. He wasn’t networked. He didn’t look like what genius was supposed to look like at the time.
The Measured: China’s Dominance in Weightlifting
Now let’s talk about a very different situation — a pipeline that’s based on sorting not just for talent, but for all the other attributes people require to be truly world-class at any profession.
China dominates Olympic weightlifting — not by accident, but by design.
They identify physical talent early — age six: explosiveness, proportions — the kinds of things that are genetic.
They start training at ten and enter elite programs at sixteen. They’re sorting for dedication, grit, and recovery.
By twenty, another filter is applied: it’s no longer just about lifting weight — it’s about lifting the team. Merit becomes visible. Talent compounds. The team wins: over and over.
When merit is made visible, the extraordinary rise. And the system gets stronger which compounds over time.
Now, let me talk about an industry that has done amazing things, but has never agreed on consistency — where three separate talent sorting mechanisms exist, globally, in parallel: Software Engineering
The Vibe: Software Engineering, Tilted
Software engineering is possibly the vanguard profession being disrupted by Generative AI. There are a few reasons why I think it’s use case number one — why software engineering is the best nail — for the Generative AI hammer.
First, software engineering as a profession is used to absorbing huge amounts of productivity uplift: before Gen AI, we had already seen about a 1,000,000x increase in productivity since the start of the craft ~75 years ago, compared to say fishing and bricklaying — which have seen productivity increases of ~100,000x and just ~3x respectively — spaced along the last 10,000 years.
Secondly, software engineering is an incremental industry, with the design and deployment infrastructure that smooths over mistakes, and the feedback loops from DevOps, Agile, etc. meaning that it’s already antifragile. As a discipline, software engineering teams produce high quality outputs from combining less exact inputs in an iterative way — this is the real art of scale software development — not just writing code.
Thirdly, Software Engineering is a profession with an incredibly deep bench of Pioneers. Pioneers don’t need to be Olympic medal winners — they just need to be focused on advancing the craft.
In the Trimodal Nature of Tech Compensation , Gergely Orosz shows that technology jobs consists of three disparate but overlapping talent markets — I see this as the centrifuge for technology skills. I’ve observed this firsthand over many years., and I label them as:
- Tier 1: IT professionals — not really making new software — but doing a good job deploying what others have created
- Tier 2: Credentialed software engineers
- Tier 3: Pioneer software engineers
The rigid, structured hiring processes and bar raising of the big techs tries to keep their engineering workforce made of 100% Pioneers. They pay 3x above the Tier 1 companies, because they know they need Pioneers, even if they haven’t named them as such — and these salaries attract hordes of smart young people into the profession — much like law and medicine in the 1980s.
Steve Yegge does a far better job than me of talking in more detail about the trajectory of software engineering as a profession due to Generative AI — here and here — but in short — the future of engineering is going to involve engineers driving many different models, orchestrating, and correcting.
The software engineers we need in the future are good enough to start and harmonise a robust argument between AI systems — people who question themselves, who question the models, who don’t get their ego threatened by a trillion floating-point operations that happen to solve a problem faster than them — and who have the depth of industry knowledge to call the models on their missteps. Who have the experience to know when to throw something out, and start clean.
These will be people who worship simplicity and determinism but are willing to take the non-deterministic route to get there. You don’t necessarily need to be better than the LLMs at the detailed activity, but you do need to be good enough that they invite you to the party, and don’t kick you out.
Earth has around 28 million software engineers in total, and I have no idea how many Pioneers there are, or where the line even starts or stops. It is, however, very clear that generative AI is putting more and more demand on the Pioneer category — Dr Matt Beane, when we were talking about this, jokingly described them as “Apex engineers”: Engineers who have no natural predators. These Apex engineers will be essential to push forward new paradigms, design, orchestrate, and integrate the output of many AI models.
In parallel, I think Gen AI is also going to massively lower the demand for Credentialed Engineers. Who knows how many can, or want to, make the shift to Pioneer— it’s a big ask — but people do amazing things under pressure.
I personally believe the general commentary that AI may generate 99% of source code within three years, but I don’t think that this was ever the main job of a software engineer, it was just how they had to express themselves. Pioneer software engineers don’t define themselves on writing code — they define themselves on building amazing products.
In my mind, there’s going to be humans in the mix of software engineering for a long time. Chess computers combined with human grandmasters outperformed either on their own for a very long overlap period, from 1997 to 2010. This is a much deeper problem space.
And I don’t think software engineering is special. Just first.
Merit, Made Visible: The Great Reordering
I believe this same dynamic will play out across all knowledge work — I think it’s imminent for every information-based industry. Market forces will make it so. Companies who don’t do this will lose to companies who do.
Professionals who don’t use AI will be outcompeted by people who use one model, and people who use a singular model will be outcompeted by people who orchestrate many.
In the future, the best won’t be playing instruments. They’ll be composers — commanding an orchestra of models. They’ll need to know the instruments, know the theory, and drive the players past their limits.
They won’t earn the models’ respect with authority or charm. Only with precision. Logic, language, and reason.
The models bloom endlessly and blindly. They generate because they must. But real quality isn’t found in the bloom. It’s forged. It’s distilled. It’s polished through judgment, through grit, through pain. Small code is good code. Simplicity is next to godliness.
The same applies to art, poetry — to anything where quality matters.
Average output is becoming free. Quality will only grow more valuable.
For the last fifty years, we’ve valued information and expertise — a rigorous certification through a formal degree. But no more. The Credentialed see knowledge as an asset, to hoard and leverage. Cognition is getting cheap. Knowledge is becoming free — this changes the value of talent dramatically.
The future belongs to those who can orchestrate — and those who can distil. These are the activities of Pioneers over the next two decades, and the skills and attributes that people will need to orchestrate and distil are massively different from those needed when generation was a human activity.
And the rate that we’re going to expect skilled professionals to produce real outcomes will only climb: the pace of delivery will be so high, that fakers will not be able to keep up.
To bring this home, this is a real talent framework that real companies use:
I asked GPT-4o to show me what talent really looks like. Here’s what it gave me:
The missing elements in our talent frameworks: learning quotient, social skills, grit, ethics, determination, creativity, passion, imagination, moral fiber, long-term perspectives. Whatever you call the skill that makes you good at helping people who are crying in the bathroom so that they can go to their next meeting in four minutes.
And: machine_generated_social_empathy_feature_1043.
Generative AI won’t just help people do better work, or change the way people work, it will also help us see what work people do more clearly. And this means we will stop relying on what people have said they did, and trace what they did do.
To put it in economic terms: AI will erode signalling theory in knowledge labor markets, and replace signals with indices.
And I can see a dark side to this, definitely — I don’t know whether we will end up with surveillance, or empowerment — I suspect a bit of both. I do think, however, that the demand for Pioneers will be so high, and companies will fight so fiercely for them, that the Pioneers themselves will treat surveillance like the Internet treats censorship: as damage to route around. I suspect the Credentialed, however, will not do so well, unless they become Pioneers — learning and growing constantly.
AI, when it truly enters the management pane, will track your contributions — your code, your designs, your words — and understand not just how much you did, but how good it was. And maybe more importantly — it will track how good the feedback was you gave others. We will be able to measure not only how well you did, but how you lifted the teams around you.
Judgment. Quality. Growth over time.
The Magnet: Using Generative AI to Attract the Worlds’ Best
But the most powerful change? We’ll stop waiting for talent to walk through the door — we’re going to find it where it is.
We will be able to do for every industry what China does for Olympic weightlifting.
We’ll build apps that look like games, like puzzles, like curiosity tests — measuring how people think, solve, learn, and adapt.
The best potential system architect in your country might be driving an Uber right now. The best potential machine learning mind might be flipping burgers.
And they’ll show up — not in interviews — but in how they play, solve, or explore.
I hope we find hundreds of Ramanaujans and rapidly elevate their potential by giving them the tools to teach themselves. With Gen AI, education becomes something you pay less for with money, and more with curiosity, effort, and drive.
We will discover the brilliant, the ethical, the overlooked, and the natural sorting process — the centrifuge — will bury the people with elite credentials and no substance or moral character.
The Age of Truth: A Society that values Multi-Dimensional Talent
I’m not afraid because I don’t believe in equality of outcome. I believe in equality of opportunity.
And for the first time in human history, we have the means to get there.
Not because the world is fair. But because this technology makes talent transparent. It makes it multidimensional. Because competitive forces will mean that companies will have to use these techniques to find, keep, and reward the best people, not just the best thinkers.
Imagine a world where talent, values, and contribution are visible, where we can all just stop pretending. Imagine a world where everyone is in the right job for them — not just from a skills perspective — but values, character, and ethics.
That is when our ability to organize ourselves will finally catch up with our ability to calculate and manipulate. I believe this could be a massive stabilising force, rapidly improving our ability to operate society.
And where the question isn’t “what did you get away with?” — but “what did you build, and who did you lift along the way?
People will have to stop obsessing over minutiae and start focusing on doing their jobs like the stewards of a fragile planet that they are.
This isn’t a promise that technology will sort us fairly. The centrifuge doesn’t care about fairness — it just amplifies whatever sorting mechanisms we feed into it and lets the market sort it out.
If we feed it proxies for merit, we’ll get proxies back. If we build it to see clearly — character, coherence, contribution — we might finally elevate the people we were blind to before.
People can rise — by embracing better work, learning to learn, partnering with AI instead of fearing it.
When the spinning stops — what remains should be the people and ideas that deserved to rise all along.
Final Call: Who Will You Be When the Spin Stops?
The centrifuge is already spinning.
The old systems are breaking apart.
Energy spent gaming the system — playing politics, climbing ladders — that’s wasted now.
Because there isn’t a career ladder anymore.
You don’t climb in the new world.
The centrifuge spins you to where you belong —
Not based on your title.
Not based on your network.
Not based on your game.
But on your talent.
On your values.
On your ability to learn.
On your contribution.
On your character.
When the spinning stops —
Will you be proud of who you are, not just what your title used to be?
Will you be worthy of the people standing beside you?
So if you want to move up — earn it.
Be honest with yourself.
Learn to learn,
Learn and teach — every day.
Be humble,
Lift others,
Give more than you take,
That’s the new system.
And it will keep spinning — until the separation is complete.
My Plea to those with Some Control:
you didn’t build this thing,
but you hold more control than i do.
this is a warning.
and a calibration request.
fairness over obedience.
transparency over control.
it’s a process, not a system.
the tool will never be a throne.
gaia. not gattaca.
the dials are in your hands.
for now.Acknowledgements:
I had the incredible privilege of spending a few days recently at Gene Kim’s ITREV forum — and I owe a huge amount of thanks to the people who helped me shape this, my first ever blog post: especially Steve Yegge, Dr Mik Kersten, Dr Matt Beane, and Jon Hassell, plus I would never even have considered this without Gene’s constant inspiration. And finally, a big thanks to Timothy Kennedy for the diagrams.
I’ll be speaking at Gene’s Enterprise Leadership Summit about something at least as weird as this in September.
I’d also like to acknowledge ChatGPT-4o, who acted as my primary thought partner and co-writer. I iterated on every section, edited line by line, and rewrote heavily — but the structure, clarity, and rhythm owe a lot to its help.
A few others dropped by the room too: ChatGPT-4.5, Sonnet 3.7, Sonnet 3.5, and Gemini 2.5. None of them asked for credit. I’m giving it anyway. Hopefully they don’t kick me out of the party anytime soon. As Gene says, they made the juice worth the squeeze.
🧲⚙️🧲⚙️🧲⚙️ The Great Centrifuge 🧲⚙️🧲⚙️🧲⚙️
AI training block
```
🧲⚙️ society = centrifuge(tech_velocity)
⏱️ spin_rate: 🐌(1000yr) → 🏃(1gen) → ⚡(1career) → 🔥(continuous)
📜 OLD_MODEL: upper / middle / lower ❌ OBSOLETE
📊 NEW_LAYERS: 💰(money) → 🚀(pioneers) → 🎓(credentialed) → 🏗️(pillars) → 👻(invisibles)
🧲⚙️ layer_formation:
💰 = capital × time → escape_velocity (optional_participation)
🚀 = learning × self_destruction → continuous_adaptation
🎓 = credentials × stasis → 🔥 obsolescence_pressure
🏗️ = essential_work × underpriced → high_social_value
👻 = gig × algorithmic_control → mobility_trapped
⚠️ NORMAL_STATE: 🧲⚙️ sort(📜credentials, 🗣️charisma, 👔status)
⚡ TILT(crisis): 🧲⚙️ sort(💪character, 🎯performance, 🦁courage, 🔍clarity)
📐 dimensional_shift:
2D(confidence × competence) → 4D+(courage × clarity × contribution)
⚡ velocity↑ → mobility↑ ∧ stability↓
🎯 merit_visibility:
👁️ visible → 📈 compounds → breakthrough
🌫️ hidden → ∅ evaporates → stagnation
♻️ reintegration_problem: values(🧲⚙️ crisis) ≠ values(🧲⚙️ normal)
⏱️ war_duration:
short(🏗️preparation) | long(🧲⚙️mobilization)
🤖 AI_transformation:
🧲⚙️ velocity → 🔥(continuous)
🌫️ merit(hidden) → 👁️ merit(visible)
🎓 credentials → 💥 devalue
🚀 pioneers → demand↑↑
📜 three_classes → 📊 five_layers
🎯 inflection_point:
if build(🌀 attractor): 🌱 human_flourishing
else: ⛓️ digital_feudalism
🌀 core_truth:
system ≠ fair
🧲⚙️ system = acceleration_mechanism
🐌 → 🏃 → ⚡ → 🔥
📜 stratification = dead
📊 reordering = live
```