You are waiting for a moment that is not coming. The rupture. The threshold. The date you can circle on a calendar when everything changes at once. Kurzweil promised it. The TED talks dramatised it. The science fiction films gave it a soundtrack.
It did not happen that way. It never was going to. What happened instead was quieter. The transformation crept in through the edges of ordinary life — through tools people were already using, workflows they had already accepted, daily decisions they had already stopped noticing they were delegating. No rupture. No threshold. A slow, continuous, barely perceptible shift in who was doing the thinking.
And most people did not notice. Because the coffee still tasted the same.
The signals that did not announce themselves
The dramatic signals got the headlines. Quantum AI. Brain-computer interfaces. Biobots. Those demanded attention. They announced themselves. The quiet version was the opposite.
- Companies started creating AI replicas of their best employees — capturing decision patterns, communication style, institutional knowledge — and deploying those replicas as autonomous agents. Not experiments. Headcount. Nobody called it the singularity. They called it digital transformation.Traditional hiring documents became irrelevant as AI assessed capability directly. Nobody called it a rupture. They called it talent strategy.
- Google’s worldwide search market share dropped below ninety percent for the first time in a decade. The entire discovery layer of the economy — how people find information, evaluate options, make decisions — was quietly rebuilt around answers instead of links. Nobody called it a revolution. They called it an algorithm update.
None of these signals, individually, felt like the singularity. That was the point. It does not feel like anything. It feels like Tuesday.
The fluency divide that is hardening into a class gap
The real divide is not between people who use AI and people who do not. It is between experienced users and newcomers. People who have been using the tools for six months or more have a measurably higher success rate — not because they are smarter, but because they have developed fluency. They iterate. They push back. They refine. They treat the AI as a thinking partner rather than a vending machine.
The returns compound. The more you use the tool, the better you get. The better you get, the more valuable it becomes. A compounding advantage that looks, from the outside, like a skills gap — but is hardening into a class gap in real time.
The divide has four rungs. Prompt literacy: knowing how to ask. Workflow redesign: rebuilding work around what the model can do. Agent orchestration: composing systems that operate without continuous supervision. Product and business-model innovation: building offers that were not possible before. Most leadership teams sit at rung one and assume they are at rung two. The compounding gap is between two and three.
The invisible infrastructure
By mid-2025, a pattern had emerged that individual news cycles could not reveal. AI was writing infrastructure code. Generating customer service responses. Drafting legal documents, screening job applicants, managing supply chains, optimising pricing, triaging medical records, approving expense reports. Each announcement looked like a feature update. A productivity improvement. A tool. Taken together, they were the quiet replacement of human judgment at every level of the organisation — one workflow at a time.
The enterprise technology graveyard filled up. Legacy systems that had run companies for decades were decommissioned not because they broke, but because they were too slow for a world where AI could do the same work in seconds. The CRM from 2019. The ERP customised in 2021. The analytics platform the board approved eighteen months ago. Archaeological artefacts of a previous era. An era that ended without anyone ringing a bell.
The professional class — knowledge workers, managers, strategists, consultants — had convinced themselves they were safe. AI would automate the routine work, they said. The creative, strategic, judgment-heavy work would remain human.
That was the story. The quiet transformation was the slow realisation that the story was wrong.
Proof of human
By early 2026, a new phrase had entered the vocabulary: proof of human. The premise was simple: in a world flooding with AI-generated content, AI agents, AI decisions, you needed a way to prove that a real human being stood behind a transaction, a conversation, a decision. Proof of human was becoming infrastructure. Think about what that means. If you need cryptographic proof that a human is involved, the default assumption has already shifted. The default is now machine. Human is the exception that requires verification.
The shift had already happened. Not in the future tense. Already. More AI agents than humans would soon be visiting corporate websites. The design, the navigation, the content architecture — all of it built for human eyes. The primary audience increasingly non-human. And nobody had redesigned anything. The infrastructure of commercial life — websites, hiring processes, customer service, procurement — was built for a world that no longer exists. The new world is already here. It just did not send a memo.
The companies most at risk feel safest
The lesson is not that AI is coming. AI is here. It has been here. It arrived without fanfare, through procurement approvals and vendor updates and workflow automations and a thousand small decisions made by people who were trying to be efficient, not revolutionary.
The companies most at risk are the ones where nothing dramatic has happened. Where the quarterly results are stable. Where the org chart looks the same as it did two years ago. Where the leadership team has discussed AI in three board meetings and concluded they are “keeping pace.” Those are the companies inside the transformation. They just do not know it yet.
The singularity happened. It is still happening. It will keep happening. And the reason most leadership teams have not noticed is that they are waiting for the dramatic version. The version from the TED talks. The version with a soundtrack. The dramatic version is not coming. The gentle version arrived through workflow updates and vendor emails and quarterly efficiency reports. It smells like coffee. It looks like a normal day at work.
Three questions your leadership team cannot afford to defer
- Where has human judgment been quietly replaced? Not the obvious places — the chatbots, the automated emails, the AI-generated reports. The subtle places. The pricing decisions nobody reviews anymore. The hiring screen that used to involve a human reader and now involves a score. Map every point where a human used to decide and a machine now decides instead. That map is the territory of the shift in your organisation.
- Who has developed AI fluency, and who has not? The gap between the people who think alongside AI and the people who use it as a search engine is already a performance gap. Within two years it will be a career gap. Within five it will be an organisational survival gap.
- What are you still building for humans? Your website. Your onboarding process. Your procurement workflow. All of it was designed for a world where humans were the primary actors. Increasingly, they are not. The redesign is not optional and it is not a future project. It is already overdue.
The only question is whether your organisation noticed.
This post is adapted from Chapter 10 of I Was 80% Wrong: What 1,400 near-daily editions taught me about strategy, sensing, and the future by Ron Immink.
