I get one question everywhere I go. Hong Kong coffee shops. Tokyo conference rooms. Zoom calls at midnight where someone is always in a timezone that hurts.
"James, will AI make it easier for ordinary people to succeed?"
I used to smile. I would say something diplomatic about new opportunities, about tools democratizing access, about the future being bright for those who adapt. Then I would change the subject before the lie fully left my mouth.
Not today. The hype has reached a fever pitch, and I have heard this exact narrative before. The mobile internet was supposed to level the playing field. Blockchain was supposed to eliminate gatekeepers. The metaverse was supposed to make geography irrelevant. Each time, the same promise: the barriers are falling, and the ordinary person is about to win.
Each time, the same reality: the barriers fell, but the winners were the same people who were already winning—just faster.
So here is the direct answer, and you can debate me if you disagree.
No. AI will not make ordinary people successful. It will make them invisible.
The "Ordinary" Trap
Let me walk through the logic slowly, because this is where most people get stuck.
When people ask about "success," they usually mean financial outperformance. They mean earning more than average, building something that compounds, creating value that the market rewards generously. And if you understand basic economics, you know that premium returns come from exactly one source: scarcity.
A top-tier engineer's ability to debug a distributed system under load is scarce. TSMC's ability to manufacture three-nanometer chips at scale is scarce. Nvidia's grip on the AI training stack is scarce. Lionel Messi's footwork is scarce. These things command premium prices because they cannot be easily replicated.
Now, what does "ordinary" mean? It means average. It means sitting comfortably in the middle of the bell curve. It means your skills, your perspective, and your output are abundant and easily replaceable.
Therefore, the phrase "ordinary person succeeding" is an economic oxymoron. It is like asking for scalding hot ice, or intersecting parallel lines. The two concepts violate each other by definition.
"But James," someone always interrupts, "AI lowers the barrier! Now anyone can write code, generate graphics, make videos. The era of the commoner has arrived!"
This is the comforting narrative that keeps people asleep. Yes, AI crashed the execution barrier. Anyone can now generate a video with a single prompt, write a Python script, or design a landing page. But here is what the narrative conveniently omits: millions of other people have the exact same tool. Your competitor has it. Your neighbor has it. The teenager in the next timezone has it.
When supply becomes effectively infinite, value drops to zero. It is a tale as old as agriculture. A bumper crop does not make the farmer rich. The oversupply drops the price of rice from fifteen dollars a sack to five. If you think you can generate a video with one prompt and make Hollywood-level money, you are ignoring the fact that millions of "ordinary" people are generating the same video at the same moment.
The people who are actually winning right now are not ordinary. They are extraordinary talents—visionaries, strategists, storytellers with mutant-level instincts—who were previously bottlenecked by capital. They could not afford a dev team, an art department, or a film crew. AI removed their resource constraint. It did not remove their need to be exceptional.
To succeed in this era, there is only one unsexy, grueling answer: you must choose to become un-ordinary.
AI Is a K-Shaped Amplifier, Not an Equalizer
I also write about AI automation and agent building. A few months back, I published a deep dive—roughly ten thousand words—on how I construct AI Skills, which are custom automated agents that handle entire workflows end-to-end. Most readers skimmed it hunting for technical shortcuts. They wanted the prompt, the tool, the one-click fix.
They missed the entire point.
AI is not an equalizer. It is a K-shaped amplifier. It is making the capable far more capable, and the lost far more lost.
Picture the letter K. The top arm shoots upward. The bottom arm drops downward. The middle—the average—shrinks.
Top-tier users are building complex, multi-agent workflows where AI handles research, drafting, coding, testing, and deployment in a single orchestrated pipeline. They are replacing entire teams with automated systems that run while they sleep. They are rocketing upward.
Meanwhile, the average user is still treating ChatGPT like a Google search bar. They ask one isolated question. They get a generic answer. They complain that "AI is useless" or "it doesn't understand me." They are spiraling downward, not because the technology failed them, but because they brought a calculator mindset to a conversation.
Here is the truth that stings: AI is a mirror. If you have a clear strategic vision and deep domain expertise, AI amplifies your capability tenfold. It becomes a force multiplier for precision. But if your goals are blurry, your thinking is shallow, and your experience is thin, AI does not fix you. It simply amplifies your noise. It generates ten times the chaos, ten times the mediocre output, ten times the confusion.
You get frustrated because the AI "isn't working." The reality is that the AI is working perfectly. It is reflecting your own lack of clarity back at you with terrifying efficiency.
The Prejudice Checklist: Your Real Moat
When I built my automated presentation-generation agent—the one that builds full slide decks from a rough brief—I did not spend most of my time telling the AI what to do. That is what beginners do. They write endless positive instructions: make it blue, use this font, add a chart here.
I spent the bulk of my energy telling the AI what not to do.
I call these negative constraints the Prejudice Checklist. They are hard-earned boundaries born from real-world failure, and they are the most valuable thing you can feed a machine.
Let me give you concrete examples, because the details matter.
Rule one: Never use a pure white background (#FFFFFF) with pure black text (#000000). To an AI, this looks sleek, minimalist, and high-contrast. Professional. But any human who has sat in a dark conference room staring at a projector knows the truth: within three minutes, that combination causes physical eye strain. The glare is exhausting. Your audience stops listening because their retinas are in pain.
Rule two: Never allow three consecutive slides with the exact same layout. The AI does not get bored. It does not experience monotony. It sees consistency and thinks, "This is clean." But human audiences are biological. Their attention drifts when visual rhythm stagnates. After three identical layouts, you have lost the room.
Rule three: Never use default chart colors without checking colorblind accessibility. The AI sees a beautiful red-green gradient and thinks it is elegant. A significant portion of your male audience sees mud.
These rules are not in the AI's training data. Or if they are, they are buried under millions of generic "best practices" that average out to mediocrity. The AI knows what usually works. It does not know what specifically fails in your domain, in your context, with your audience.
This is your ultimate competitive moat. AI models are trained on the statistical average of human knowledge. They know the positive, generic rules—the safe paths, the consensus answers. But they do not have your failure data. They do not know the painful, expensive mistakes that keep you awake at night. They do not know the client who walked out because the presentation was too dense, the campaign that flopped because the tone was slightly off, the product that died because the onboarding flow assumed too much prior knowledge.
Your value is no longer in executing the task. Your value is in the hyper-specific, hard-earned boundaries you set for the machine.
A one-click "AI Skill Generator" will give you a hollow shell. It will follow the steps, check the boxes, and produce something that looks correct. But it will lack the soul of your failure data. It is like a recipe written by someone who has never tasted food. It hits all the technical marks and misses the human experience entirely.
The masters are not the ones who prompt best. They are the ones who know exactly what to forbid.
The Interaction Mindset: Stop Treating AI Like a Calculator
The final hurdle is not technical. It is psychological. Most people still use what I call the Legacy Software Mindset with AI.
When you use Excel, you input a formula and get a definitive result. One action, one output. The relationship is transactional and final. You expect correctness on the first try.
AI does not work this way. It is conversational. Iterative. Probabilistic. If you ask an AI to design a strategy and the first draft is mediocre, the ordinary user gives up and declares, "AI is useless." They treat the first output like a final exam grade.
The elite users treat AI like an incredibly fast, slightly naive junior colleague who never sleeps and never takes offense. They say: "This is too dense. Condense page three. Change the tone to be more authoritative. Add a focus on E-E-A-T principles. Now rewrite the conclusion to match the new opening."
To bridge this gap, I use two mandatory techniques in my daily workflow. They are not optional. They are the difference between getting garbage and getting gold.
Technique one: The Reverse Interview. When I prompt an AI, I always add this at the end: "The information I have given you is incomplete. Before you execute this task, ask me the five most critical questions you need answered to fully understand my goal."
This forces the AI to extract context from my brain instead of hallucinating based on generic patterns. It turns a one-way command into a two-way conversation. The AI surfaces gaps I did not know I had. It asks about the audience, the constraints, the success criteria, the failure modes. It forces me to think before it executes.
Technique two: The Blueprint Protocol. I never let an AI generate final output on the first pass. Never. I force it to output a project plan, an architecture document, or a detailed outline first. I review it. I challenge it. I adjust the scope, the sequence, the emphasis. Only after I approve the blueprint do I allow the AI to generate the final code, content, or design.
This prevents the sunk-cost trap. Once you see a polished draft, you become attached to it. You start editing instead of restructuring. The Blueprint Protocol keeps you in the architect's seat, not the editor's chair.
Build the Un-Ordinary Business
At Mercury Technology Solution, I see this K-shaped divergence play out in real time. That is why we do not hand our clients generic AI tools and wish them luck. We build systems that encode expert prejudices into the infrastructure.
The Mercury Business Operation Suite captures real operational data—sales pipelines, procurement patterns, inventory flows—so that any AI working with the business has context, not just capability. It knows your seasonality, your supplier delays, your customer churn triggers.
Mercury Muses AI handles the repetitive execution—the drafting, the scheduling, the data processing—so human experts can focus on defining boundaries and strategy. It is not a magic wand. It is a skilled laborer that works under expert supervision.
Through our Generative AI Optimization frameworks, we do not just ask AI for traffic. We architect content specifically designed to feed algorithms exactly what they need to see—E-E-A-T signals, structured entity relationships, citable semantic blocks—while avoiding the structural pitfalls that bury ordinary websites. We build the Trust Layer that outlasts any single model or platform shift.
The gap between humans is widening. AI will not save the ordinary worker or the average business. It will only accelerate the masters.
So stop looking for the easy way out. Stop waiting for the AI fairy godmother to drop a pie on your head because you learned how to type a prompt.
Build your failure data. Encode your prejudices. Master the interaction mindset.
Choose to be extraordinary.
Stay ahead of the curve.
— James
