TL;DR: A lot of professionals are currently panicking because they realize "Digital Employees" (AI agents) can do their jobs. What they don't understand is that this architectural shift has been happening for 25 years. I have been building "digital employees" since I was selling video game macro-bots in college. If your entire career strategy is to be a low-level executor—a person who simply assembles the Lego blocks someone else designed—you have never been safe. You have always been a replaceable gear in a system designed to eliminate you. The endgame of the AI era is mathematically predetermined. You only have one choice: Do you align yourself with the system's architecture, or do you bet against the inevitable?
James here, CEO of Mercury Technology Solutions. Hong Kong — April 8, 2026
Whenever I discuss the reality of AI replacing human labor, people become incredibly emotional. They talk about fairness, loyalty, and the intrinsic value of human effort.
I always give them a very simple piece of advice: Never bet against the physics of a system. If you place a piece of sugar on the ground, betting that the ants will eventually carry it away is a guaranteed win. Betting that the sugar will sit there untouched forever is financial suicide. The universe operates on entropy and efficiency. Emotion has no place in systemic architecture. The system simply observes.
Right now, white-collar workers are terrified that AI tokens are replacing their salaries. But the sky isn't suddenly falling; the sky has always been falling. You just weren't paying attention.
1. The Digital Employee is 30 Years Old
In the early 2000s, while I was in university, I wrote and sold macro-scripts (bots) for online MMORPGs.
Back then, "gold farmers" would manually grind in a video game for 30 days straight just to make $70USD a month. I looked at that system and realized: Why dig for gold when you can sell the excavator? I went to the gold farmers and sold them a script that automated their entire month of labor.
That was the prototype of the "Digital Employee." Back then, it burned prepaid game cards. Today, it burns API tokens. The underlying architecture is exactly the same.
Even back then, rudimentary "Text Generators" existed for online novelists. You could input a faction name, a character profile, and a setting, and it would spit out generic chapters. If you think generative AI started with ChatGPT in 2022, you are completely blind to the historical timeline of automation.
2. The Architecture of Disposability (The Lego Trap)
The most brutal example of this was in the software engineering industry around 2009.
I was working as an architect in a mega software company. We had to adapt our software (and test) to run across hundreds of different hardware chips, operating systems, and client applications. If we had solved this by just hiring more junior developers to write customized code for every client, the payroll would have bankrupted the company.
So, what did we do? We built a workflow.
We created a core Architecture Department with just four elite engineers. Those four architects wrote the "Virtual Functions"—the high-level abstractions that covered the entire system. We then packaged these functions into simple, idiot-proof APIs (Application Programming Interfaces).
We turned the complex system into Lego blocks.
The hundreds of junior "Application Engineers" and "Software Engineers" we hired below us didn't need to understand how the system worked. They just needed to know how to snap the Lego blocks together based on the client's request. Just like manufacturing line labour.
Those junior engineers were effectively human meat-puppets for the four architects. We didn't need to train them. We didn't need to promote them. If they quit after three years because they realized they had no career path, we just hired a new batch of cheap graduates from $1200RMB to snap the Legos together.
I always get asked: "What are the opportunities for ordinary people?"
Look at those junior engineers. Did they have an opportunity? They were isolated from the core architecture. They spent their days tweaking Lego blocks. They had zero chance of survival. Getting fired at age 35 wasn't a possibility; it was a mathematical certainty.
If you are just executing templates someone else designed, you are not a professional. You are a biological placeholder waiting for an algorithm to become cheap enough to replace you.
3. The Endgame: Platform vs. Creator
Back in my college days, I tried to sell that primitive Text Generator to an early online novelist whose work I admired.
His response was a masterclass in systemic foresight. He told me: "If a software can truly generate top-tier novels, what is the endgame of this industry? The endgame is that there will be no human writers. The platform (the publisher) will just generate the novels themselves and keep 100% of the profits. Why would they let me use the tool to make money off them? Conversely, if the tool produces garbage, then the platform's strategy will simply be 'Gu (蠱) Poison warfare'—they will force thousands of human writers to compete ruthlessly until one genius survives, and the platform will harvest their success."
He understood the terminal architecture of his industry. He knew that adopting the tool wouldn't save him if the tool became omnipotent.
Conclusion: Pick Up the Wallet
The future is already written. 18 years ago, I asked my junior engineers: "Is your career strategy really to just pray you stay cheaper than a fresh college graduate forever?"
Today, I ask you the exact same question: Is your strategy for 2026 really to hope you remain cheaper than a monthly Claude API subscription? If you see a wallet lying on the street, you pick it up. If you don't pick it up, your competitor will. And when they use that capital to destroy your business, you will regret your hesitation.
You cannot fight the architecture of the system. You either elevate yourself to the Architect level, or you accept your fate as a Lego block. Align yourself with the inevitable.
Mercury Technology Solutions: Accelerate Digitality.
