The F-18 on the Runway
Moore's Law described the doubling of computing power every two years. It has been running since 1965 — the longest sustained rate of technological progress in modern history. It built Silicon Valley. It shaped every business playbook written in the last sixty years.
As of this decade, it is no longer the fastest thing on the road.
Picture a Sunday driver on the Great Ocean Road. Top down. Leisurely pace. For sixty years he has been the quickest thing any of us had ever seen. Now picture an F-18 rolling onto a runway alongside him — and taking off.
That is the moment we are in.
The technology doubling behind AI is no longer exponential. It is hyperbolic — a curve so steep it bends the frame it is drawn in. Capability gains that took Moore's Law two years now arrive in weeks. What was a single research paper last summer is a commoditised API this quarter and, by the time the board reconvenes, an entire business model.
If the feeling in your business right now is that the ground keeps shifting faster than you can get your footing — that is not a perception problem. That is the ground shifting faster than you can get your footing.
We have entered a new economic phase. Not a new product cycle. Not a new vendor category. A new phase of human economic development, qualitatively different from the one that preceded it.
It needs a name. The polite language of "AI transformation" does not do it justice — AI is just the visible artefact. The phase itself is something bigger.
We are calling it Phase Three. And if you are running a business today, the most important thing to understand is this:
Your playbook was not written for it.
Three Phases
To see why, it helps to look at the two phases that came before.
Phase One — Diminishing Returns. The 400-year-old operating system. Linear, effort-intensive growth. Output is proportional to input. If you want more, you do more. Plant more fields. Hire more hands. Work longer hours. Most of human economic history has run on this logic, and most of the management orthodoxy still taught at business schools assumes it.
In Phase One, growth curves plateau. The tenth unit of effort produces less than the first. This is the world of candlelight and clock towers, of agricultural fortunes and industrial hierarchies. It is slow and it is honest. You get what you earn.
Phase Two — Exponential Growth. The Moore's Law era. From roughly 1965 to the early 2020s, silicon-based technology unlocked the first genuinely exponential growth curve in human economic history. A billion doublings over ninety-nine years. From punch cards to smartphones. From the PDP-1 to the iPhone. From the fax machine to Stripe.
Phase Two did not replace Phase One. It ran alongside it. But for the first time, there were companies whose growth defied the old rules — whose unit economics improved with scale instead of deteriorating. These companies, the digital natives, came to dominate every category they entered. And every established business built its defensive playbook around trying to become one of them, or at least to fend them off.
Then the curve bent again.
Phase Three — Diminishing Effort. Recursive AI. The phase we are living through right now. For the first time in economic history, capability improvements compound without a corresponding increase in human input. The input side of the equation is shrinking. Equivalent or greater output with progressively less human effort.
A billion doublings in under six years — the projection for the next round of AI capability gains. For comparison, Phase Two delivered a billion doublings in ninety-nine.
This is not Phase Two moving faster. This is a different curve.
Why Phase Three Breaks the Rules
The temptation — particularly if you have built a career through Phase Two — is to treat Phase Three as a speed problem. "We need to adopt AI faster." "We need to digitise more aggressively." "We need to retrain our workforce." All true. None of it sufficient.
Exponential and hyperbolic curves are not the same shape. That distinction matters.
An exponential curve doubles at a constant rate. It is predictable, if relentless. You can plan against it. You can build an org chart that anticipates the next doubling, hire against the next one after that, and budget for the one after that. A thousand management consultants built entire practices helping companies do exactly this through Phase Two. It worked.
A hyperbolic curve is different. The doubling time itself shrinks. It doubles, then doubles faster, then doubles faster again, until the curve goes very nearly vertical. At that point the planning horizon collapses. You cannot budget for what happens in six months because the thing that is happening in six months will have happened in six weeks.
The mathematics of Phase Three removes a hidden assumption that every Phase Two playbook relied on: that human capacity was the bottleneck. In Phase Two, you were bottlenecked by how fast you could hire, how fast you could train, how fast you could build. Your moat was the speed at which you could scale human effort.
Recursive AI removes that bottleneck. A four-person company can now do what a four-hundred-person company did five years ago, across a widening range of knowledge work. Not because AI replaces those four hundred people. Because the remaining four, directing well, compound their output in ways the four hundred never could.
Phase Three is the phase in which human effort is no longer the rate limiter. Judgement is. Direction is. Meaning-making is. The quality of thinking at the top of a business — and the quality of the thinking it invites from everyone else — becomes the competitive substrate.
And that is a very different game.
What Your Playbook Gets Wrong
Concretely, here is what breaks.
The hiring playbook. The Phase Two assumption was that talent was scarce, so you hired for skills and trained for fit. In Phase Three, the skills commoditise within a quarter of being hired for. Hiring for skills is hiring for a position that will not exist by the time the person ramps. What remains scarce is judgement — the ability to direct a capability-compounding system toward outcomes that matter.
The competitive moat playbook. Phase Two moats were built on scale, data, and distribution. All three are being eroded by the speed at which Phase Three competitors can stand up equivalents. A two-person startup can ingest your entire product category, build a competitive feature, and launch into your market in the time it takes your board to approve a response. Moats built on the slowness of competitors no longer work, because competitors are no longer slow.
The growth model playbook. Phase Two growth was unit-economics driven. Input cost, output value, margin compounds with scale. Phase Three growth is judgement-economics driven. The marginal cost of execution is approaching zero. What scales is the quality of the decisions being made, and the coherence of the organisation making them. Most growth models do not have a column for that.
The org design playbook. Phase Two org design assumed you would keep adding people to scale output. Phase Three org design assumes the opposite. The question stops being "how many people do we need?" and becomes "how good is the thinking of the people we have, and how clearly are they directing the systems they operate?" Hierarchy exists to coordinate effort. If effort is no longer the bottleneck, most of the hierarchy loses its justification.
The leadership playbook. This is the one that hurts most. Phase Two leadership was about running the machine. Phase Three leadership is about upgrading the humans directing the machine. If you are still running Phase Two leadership, you are optimising the wrong variable.
The Window
A common response to this argument, particularly from those who survived the Phase Two transition, is: we have been here before. Every generation thinks its technological moment is the defining one. Internet. Mobile. Cloud. Each felt world-changing. Each settled into a new normal.
The difference this time is the adaptation window.
The Phase One to Phase Two transition took roughly a century. From the first commercial electric motor in the 1870s to the point at which electrification-native business models unambiguously dominated the economy, you can count the years in generations. Firms that entered the transition unprepared had decades to adapt, and many did. The children of the people who built Phase One companies built the Phase Two ones.
The Phase Two to Phase Three transition is not taking a century. On current trajectories, the meaningful adaptation window is two to three years.
This is not a prediction from a futurist. It is arithmetic. If capability is doubling in weeks instead of years, the gap between a business that started adapting in 2024 and one that starts in 2027 is not three years of catching up. It is a gap that cannot be closed inside a business cycle. By the time the 2027 starter has a coherent Phase Three strategy, the 2024 starter has iterated on it for twenty-four doublings.
Australian businesses are disproportionately exposed. Twelve per cent of Australian leaders told Deloitte in 2026 that GenAI is transforming their business. The global average is twenty-five. The firms above that Australian line will define the next decade. The firms below it will be acquired or outcompeted inside it.
The adaptation window is closing now, not later.
The Upgrade Phase Three Demands
So what upgrades?
Not the technology. The technology will upgrade itself. The compounding capability gains of Phase Three are the definition of technology that upgrades itself — that is the whole point of the phase.
What has to upgrade is the intelligence directing it.
Human cognition has, in every prior phase, been the bottleneck and the differentiator simultaneously. IQ — the raw processing power of the human mind — built Phase Two. The smartest engineers, the sharpest strategists, the quickest analysts. Then came EQ — emotional intelligence. The recognition that human systems run on relationships as much as logic, and that leaders who could read the room outperformed leaders who could only read the spreadsheet. EQ became table stakes. A leader today who lacks emotional intelligence is not a leader, regardless of IQ.
Phase Three commoditises the first and assumes the second. The marginal value of raw IQ drops when any organisation can summon a functionally-infinite supply of analytical capability on demand. The marginal value of EQ stays roughly constant — humans still need to work with humans.
What emerges as the new differentiator is a capacity that has always existed but has never before been commercially critical: Spiritual Intelligence, or SQ. The capacity for meaning-making, for purpose alignment, for systemic awareness, for integrative cognition that can hold the full picture of an organisation, a market, and a civilisation simultaneously.
This is not soft language. It is the hardest, most commercially important cognitive upgrade available to a leader right now. SQ is how you decide what the machines should do. SQ is how you direct compounding capability toward outcomes that build, rather than outcomes that hollow out. SQ is how you tell the difference, inside your own business, between acceleration and collapse.
In a phase where anyone can do anything, what separates the companies that thrive is the quality of the thinking that decides what should be done.
That capacity is SQ. And almost no Phase Two leadership development programme has trained for it.
What Phase Three Businesses Do Differently
The firms beginning to adapt coherently to Phase Three share a small number of visible traits. None of them is a technology choice.
They treat AI capability as capacity, not headcount. Instead of asking "how many AI tools do we need to deploy?" they ask "what could a team of three now do that a team of thirty did last year?" The conversation starts with outcomes and works backwards to structure.
They invest in leader upgrade before workforce upgrade. The logic is unforgiving. If the leadership is still Phase Two, no amount of workforce training recovers the gap. If the leadership is Phase Three-capable, the workforce follows. These firms are investing disproportionately in senior judgement, systemic thinking, and reflective capacity.
They collapse the planning horizon. Annual operating plans are replaced with quarterly ones. Quarterly are replaced with six-weekly. Strategic commitments are made with explicit review points, not fixed end dates. The old rigidity was a feature of Phase Two's relative predictability. Phase Three demands commitment with built-in reassessment.
They build thin, not wide. Phase Two growth was adding scope. Phase Three growth is removing it. These firms get ruthlessly clear on what they are for, what they are not for, and where the capability stack serves the purpose rather than the other way around.
They pay serious attention to why. Purpose and values in Phase Two were brand language. In Phase Three they are operating infrastructure, because when marginal cost of execution approaches zero, the only thing that stops a company from doing anything is the clarity of what it should do. Purpose stops being a page on the website and starts being the daily decision filter.
They choose fewer, deeper partnerships. Not because consolidation is fashionable, but because the pace of change makes shallow relationships non-viable. You cannot coordinate Phase Three capability gains across a dozen vendors each trying to sell you their own. The firms that win are the ones with embedded partners moving at their speed.
None of this is a technology purchase. All of it is a leadership decision.
The Question in Front of You
Your business playbook was written in Phase One. It was rewritten for Phase Two. The question in front of you now is simpler than any framework makes it sound.
When you look at your business in eighteen months' time, is what you see today going to be recognisable? And if the honest answer is no — what are you doing, today, to be the version of your company that is doing the unrecognising, rather than the version being unrecognised?
Phase Three is not a prediction. It is a description of where we already are. The only question is whether the leadership sitting at the top of your business is built for it.
If you are not sure, that is a signal worth listening to. The window to answer it is not long.
— Thomas W. Green, April 2026