TL;DR: Commercial pilots are paid massive salaries not just because they are smart, but because they operate multi-million-dollar machines. In 2026, paid AI models and autonomous Agents are the new commercial jets. If you are still using free AI, you are throwing paper airplanes while the rest of the industry is flying Boeing 777s. As a CEO, I look at labor entirely through the lens of "Token Economics." I would much rather spend $20,000 a month burning API tokens than $500,000 a year on a middle-management team, because tokens do not sleep, complain, or require HR mediation. You need to start accumulating your AI "flight hours" immediately, or you will be grounded permanently.
James here, CEO of Mercury Technology Solutions.
Hong Kong — March 29, 2026
A founder friend of mine recently told me a hard truth about his company’s 2026 budget: "Half of my sales team doesn't want to use AI. So, I'm only buying paid AI licenses for the half that does. The other half will be gone by Q4."
Since the explosion of autonomous AI Agents early this year, the computer has fundamentally upgraded to version 2.0. If you are still relying on the free tiers of ChatGPT or Claude, you are functionally illiterate in the new economy. Free AI cannot fly.
To survive the next 12 months, you need to understand exactly how the people writing your paychecks—the CEOs and business owners—are currently calculating the value of your labor.
1. The "Flight Hours" Mandate
Right now, every CEO is making two critical resource allocation decisions:
- Who gets the $20/month license? This is the baseline. Not everyone will get one. If you get it, you are considered trainable. But make no mistake: the $20 "Plus" account is just a trial. It lets you feel what it's like to leave the runway.
- Who gets the $200/month operational budget? Every department needs one heavy operator—a pilot who actually uses the advanced agents to complete massive, end-to-end workflows.
If I am the boss, who do I give that $200 account to? I give it to the person with the most "flight hours." I give it to the employee who has already proven they know how to steer the machine without crashing it.
If you want to secure your job, you need to be the person burning those tokens. Because the standard interview question for any white-collar job in 2026 is no longer, "What are your strengths?" It is: "What is your monthly AI token burn, and what structural outcomes do you generate with it?"
2. The Cold Math of Token Economics
When Jensen Huang recently suggested that professionals should be burning at least half their salary in AI tokens, a lot of employees thought he was crazy. Why would a company spend $20,000 a month just on API calls?
Let me explain exactly how a CEO views Token Economics.
Let's say you are a mid-level manager at a bank here in Central, Hong Kong. You manage a team of 6 analysts.
- Average salary: HK$70,000/month.
- Annual cost (with a 13th month): HK$5.46 Million per year.
That HK$5.46 Million buys me human labor. Human labor comes with massive friction. It comes with sick days, office politics, emotional burnout, 40-hour weekly limits, and inevitable human error.
Now, look at the alternative. If I take a fraction of that budget—say, $20,000 USD (roughly HK$156,000) per month—and spend it entirely on API tokens to power a fleet of autonomous AI agents, what do I get?
- Zero Friction: Tokens execute code 24/7/365.
- Infinite Scalability: If volume doubles, I don't need to spend 3 months interviewing new hires; I just increase the API limit limit.
- Absolute Predictability: The agent does not get tired at 4:00 PM on a Friday and make a data-entry mistake.
To an employee, spending $20,000 a month on "invisible computer credits" sounds insane. To a business owner, replacing a $5.46M payroll with a highly predictable, frictionless token burn rate is the most beautiful financial equation on earth.
3. The Transition Phase
We are not at the point where humans are entirely obsolete. The AI still needs a pilot.
But the ratio is violently shifting. I no longer need 6 analysts. I need one elite pilot (the manager) armed with a massive token budget to execute the work of 60 people.
The money I save on the 5 analysts I let go is immediately reallocated into the API token budget to make the single pilot even more powerful. Tokens are simply cheaper, faster, and more reliable than salaries.
Conclusion: Become the Pilot
Do not complain that the company is spending more on software than on bonuses. Recognize the shift in the economic winds.
If you are currently executing tasks that do not require an AI, you are the equivalent of a horse-drawn carriage driver watching the first automobiles roll off the assembly line. Your goal today is to grab the keys to the new machine. Pay the $20 yourself if you have to. Rack up the flight hours.
When your boss eventually comes looking for someone to hand a $2,000 or $20,000 monthly token budget to, you must be the only logical choice in the room.
Mercury Technology Solutions: Accelerate Digitality.
