Every technology follows an abstraction curve
In January 2023, ChatGPT passed a Wharton MBA exam. Like everyone else, I got distracted by the headline. The obvious implications were for education, white-collar work and professional knowledge. If a machine could pass an MBA exam, what did that say about business schools, consultants, analysts, managers and knowledge workers?
But I missed the deeper signal for months. The important point was not that a machine could pass an exam. The important point was that it passed using plain English. No code. No technical skill. No specialist interface. A person typed a question, and the machine answered at MBA level.
That is not only an AI story. It is an abstraction story.
The rule most leaders miss
Every technology follows the same curve. It starts complex, expensive and controlled by specialists. Then it gets simplified, packaged, hidden behind interfaces and made usable by more people. The machinery does not disappear. The complexity is pushed down into the system so the user no longer has to deal with it directly.
Electricity started with engineers wiring generators. Then the interface became a plug in the wall. Photography started with chemicals, plates and darkrooms. Then the interface became a button on a phone. Computing went from punch cards to assembly language, from high-level languages to apps, and now from apps to natural language prompts.
Each layer moved the technology further away from the specialist and closer to the user. That is what abstraction does. It removes friction. It hides complexity. It turns expertise into an interface.
But there is a harder business implication. Abstraction does not just democratise creation. It destroys scarcity. If your margin depends on specialist knowledge, technical difficulty, process complexity or access to tools others cannot use, abstraction is a direct threat to your business model.
From code to English in thirty years
In the 1990s, building software required professional programmers. The bottleneck was engineering talent. In the 2000s, frameworks reduced the effort. Programmers were still needed, but they could build faster. In the 2010s, low-code and no-code platforms arrived. Business analysts and process owners could build workflows without writing traditional code.
I worked with AgilePoint during that period. They had been building codeless architecture since 2003, years before the market caught up. They understood something the industry resisted for another decade: code equals choke. Every line of code can become a future bottleneck. Every dependency slows change. Every technical layer can become tomorrow’s rigidity.
Then AI entered the stack. The barrier dropped again. The person building no longer needed to understand every step of the logic. They needed to understand the desired outcome. In 2023, the next abstraction layer became visible: natural language.
“The hottest new programming language is English.” I wrote that in March 2025, but the shift had started earlier. Type what you want. The machine drafts, builds, reasons, tests and explains. That is not an incremental improvement. It is the compression of a thirty-year abstraction curve into a conversational interface.
It applies beyond software
This pattern is not limited to code. Biology is moving along the same path. Gene editing started in specialist labs. CRISPR lowered the barrier. AI-assisted biology lowers it again. The direction is clear: biology is becoming more programmable, more modelled and more open to instruction.
Robotics is following the same curve. Once it required factories, engineers and controlled environments. Now the interface is moving toward instruction, demonstration and adaptation. The specialist knowledge still matters, but more of it is being pushed into tools, platforms and models.
Even meditation has been abstracted. Ancient practice became a ten-minute app, a wearable signal and a guided breathing protocol. Something that once required discipline, teaching and long practice has been packaged into an interface.
The same pattern keeps repeating. Complexity gets hidden. Access expands. The interface becomes simpler. The specialist loses monopoly power.
The strategic consequence
The mistake is to think abstraction only creates opportunity. It also creates competition. When the interface becomes simpler, more people can enter the game. Your next rival may not be a well-funded company with fifty engineers. It may be one person with a no-code stack, AI agents, outsourced infrastructure and a sharper understanding of the abstraction curve than you.
This is why leaders need to stop treating AI, low-code, robotics, CRISPR and automation as separate trends. They are expressions of the same underlying movement. Technology is becoming easier to instruct. Capability is being unbundled from expertise. Scarcity is being attacked at the interface.
The question is not whether your sector will be affected. The question is which part of your value proposition still depends on friction.
The leadership question
What is the core technical skill your business currently depends on — the thing that creates competitive advantage through the scarcity of talent? Now ask: where is that skill on the abstraction curve? And what happens to your business model when it reaches the endpoint?
The technology is becoming LEGO. The question is who learns to play first.
This post is adapted from Lesson 2 of my upcoming “Was 80% Wrong: What 1,400 near-daily editions taught me about strategy, sensing, and the future “. DM me if you want a review copy.
