In March 2021 I sat on nationwide tv and stated we had been on the precipice of a quantum shift.
AI was eradicating task-based roles. The organisations that will survive had been those with leaders who had already realized to ship leads to chaos and constraint.
The interviewer moved on. The phase ended. The world saved going.
I had began my enterprise 5 years earlier on a single conviction: that the operators constructing startups from nothing had been a distinct calibre of chief fully, and that the situations they labored in weren’t a startup quirk however a preview of how each organisation would ultimately need to function.
Final week, that prediction arrived. Not step by step. Unexpectedly.
5 days that modified the dialog
On Tuesday, WiseTech Global, the Sydney-based logistics software company, announced it was cutting 2,000 jobs, practically a 3rd of its international workforce throughout 40 international locations, as a part of a two-year AI-driven restructure. Some groups, together with its not too long ago acquired US cloud division E2open, face cuts of as much as 50%.
The identical day, Commonwealth Financial institution confirmed 300 know-how roles had been being eradicated. WiseTech’s inventory rose 11%. CBA’s buyers barely flinched.
On Thursday, Block announced it was reducing its workforce from just over 10,000 to under 6,000. Greater than 4,000 individuals advised to go away. Fill up 24%.
In his letter to shareholders, Block CEO Jack Dorsey was unambiguous in regards to the cause. “Intelligence instruments have modified what it means to construct and run an organization. We’re already seeing it internally. A considerably smaller crew, utilizing the instruments we’re constructing, can do extra and do it higher.” He was equally direct about why he acted decisively quite than step by step: “Repeated rounds of cuts are harmful to morale, to focus, and to the belief that prospects and shareholders place in our potential to guide.”
WiseTech CEO Zubin Appoo was simply as plain. “The period of manually writing code because the core act of engineering is over.”
These should not euphemisms. They’re executives stating, on the document, that their earlier headcount was a legacy of how organisations used to need to function, and that AI has made that mannequin out of date. The market agreed, loudly, each instances.
The Substack put up that broke the SaaS market
Earlier than Block and WiseTech made headlines, one thing else occurred that most individuals outdoors monetary markets missed.
Over the weekend of 22-23 February, a 5,000-word Substack put up by Citrini Analysis went viral.
Titled “The 2028 International Intelligence Disaster,” it was written as a fictional retrospective from two years sooner or later, laying out a state of affairs during which AI adoption triggers a cascade of white-collar job losses that ultimately crashes the S&P 500 by 38%.
Co-authored by James van Geelen and Alap Shah of Lotus Expertise Administration, it described itself as a thought train, not a prediction. The market handled it as one thing nearer to a analysis.
On Monday 23 February, enterprise software program shares had been obliterated. Salesforce, ServiceNow, Adobe, and Workday every dropped 7% or extra. Intuit fell practically 11%.
The Dow Jones shed 800 factors in a single session. The iShares software program ETF hit a brand new 52-week low, down practically 30% 12 months so far. Michael Burry amplified the put up on X with three phrases: “And also you assume I’m bearish.”
The mechanism Citrini described is value understanding as a result of it’s not theoretical. It’s already taking part in out. AI reduces human headcount. Fewer people means fewer seat licences. Fewer seat licences means SaaS income contracts. SaaS corporations minimize their very own workers to defend margin.
That compounds the white-collar job losses throughout the broader financial system. The businesses that constructed the software program financial system are being disrupted by the identical drive that’s disrupting their prospects.
WiseTech’s CEO confirmed the loop is actual when he introduced they had been abandoning seat-based pricing fully, transferring to a transaction mannequin in order that their income is tied to throughput, not headcount.
“For SaaS companies that monetise on seats or customers,” he stated, “AI will disrupt them.”
He stated the quiet half out loud, on an earnings name, in entrance of buyers. The share value went up.
What final week truly means
Taken individually, every of those occasions may be defined away. Block had over-hired throughout the pandemic years. WiseTech had particular structural pressures. CBA is a financial institution making incremental effectivity selections. The SaaS selloff was partly pushed by a fictional state of affairs in a finance publication.
Taken collectively, they describe one thing that’s not potential to dismiss as a collection of remoted incidents.
US corporations introduced 108,000 layoffs in January 2026 alone, up 118% from a 12 months earlier, the best determine for any January since 2009. On the present price, 2026 is on monitor to provide 270,000 tech job losses, outpacing 2025 considerably. And in contrast to the post-pandemic correction of 2023 and 2024, which was primarily about reversing over-hiring, this wave is structurally completely different.
Organisations should not chopping to get again to regular. They’re chopping to function in a manner that was not beforehand potential.
The share value information makes this distinction clearly. When an organization cuts workers as a result of it’s in monetary misery, the market punishes it.
When it cuts as a result of AI is enabling it to provide the identical or higher output with fewer individuals, the market rewards it.
Block was not in misery. Its gross revenue grew 24% within the quarter it introduced the layoffs. WiseTech reported a first-half revenue 6% forward of market consensus on the identical day it introduced 2,000 redundancies.
These should not corporations retreating. They’re corporations accelerating by eradicating what AI has made redundant.
The last decade that constructed towards this second
To grasp the place we’re going, you must perceive what the final ten years had been producing.
Whereas the company world ran innovation labs and debated digital transformation roadmaps, the startup ecosystem was working a reside experiment in what organisations seem like if you strip away each layer of course of, each committee, each function that exists as a result of it has all the time existed.
Startup founders had no selection however to function this manner. You rent for outcomes since you can’t afford to rent for credentials and discover out later the outcomes should not there.
You make selections quick as a result of sluggish selections are deadly. You run lean not as a philosophy however as a survival situation.
You change into AI-native not as a result of it’s a strategic precedence however as a result of it’s the solely technique to shut the useful resource hole between a twenty-person firm and an incumbent with a thousand.
That atmosphere, compounding over a decade, produced a selected type of government. Somebody who can run a perform and not using a crew behind them.
Who makes irreversible calls with incomplete info. Who doesn’t want situations to stabilise earlier than they will carry out. Who treats the absence of course of as a bonus to be exploited quite than an issue to be solved.
The company world largely dismissed all of this because the chaotic behaviour of corporations which may not survive. Some didn’t.
However the ones that did produced a era of operators that each restructuring organisation now wants and most do not know the best way to discover.
The query no person is answering loudly sufficient
The restructuring choice is simple. A board could make that decision in a day.
The onerous query is what comes subsequent. Once you take headcount from a thousand to 5 hundred, if you collapse three features into one, if you rebuild round AI as infrastructure quite than AI as a function, the individuals who stay must function at a degree that almost all of them have by no means been requested to achieve.
They should make selections that committees used to make. Lead groups at a tempo that enormous organisations had been by no means designed to maneuver at. Ship outcomes with out the methods and buildings that their whole careers had been constructed round.
AI doesn’t remove the necessity for distinctive leaders. It eliminates the buffer that common leaders used to cover behind.
The layers of course of, the big groups, the sluggish cycles that saved organisations working regardless of mediocre management on the high. All of that’s being stripped away.
What stays is a direct line between the standard of the individual on the high and the efficiency of the organisation beneath them.
In a restructured organisation with no redundancy in-built, that line isn’t a margin query. It’s an existential one.
What the following decade appears to be like like
The final decade was the startup ecosystem proving a distinct working mannequin was potential. The subsequent decade is each different sector being pressured to undertake it.
The organisations that thrive won’t be those with probably the most refined AI stack. These instruments are a commodity now. Each competitor has entry to the identical fashions, the identical infrastructure, the identical functionality.
The differentiator is the human who is aware of the best way to use it: who has already constructed within the situations that AI restructuring creates, who has the instincts that come solely from having operated and not using a security internet, who doesn’t want a playbook as a result of they wrote the final one themselves.
That individual exists. They’ve been solid by a decade of startup situations that the company world spent ten years both ignoring or dismissing as too dangerous.
In 2016 I believed that the individuals constructing startups had been constructing the way forward for management, not only for startups however for each organisation that will ultimately need to function the way in which startups already did. In 2021 I stated the shift was coming. This week, throughout 5 days in February 2026, the numbers arrived.
Block. WiseTech. Commonwealth Financial institution. A Substack put up that moved markets. A decade of slow-motion change, after which suddenly.
The subsequent ten years will probably be outlined by which organisations perceive what these numbers had been truly saying and transfer quick sufficient to behave on it.
- Dexter Cousins is the founding father of Tier One People, Australia’s main government search agency for fintech.
