There is a phrase making the rounds in Gen Z and Gen Alpha tech circles right now: "Delete everyone born before 1990 from your contacts. Don't let their boomer energy drag you down. Just talk to AI."
The premise is tempting. Why deal with the friction, ego, and fatigue of older human mentors when you have access to an all-knowing, infinitely patient Large Language Model? Can you really unlock the hidden depths of the business world using just You + AI?
The answer is: it depends entirely on who you are.
Let me debug this mindset and show you the stark difference between using AI as an executive versus using it as a novice.
The Boss's Utopia: Why Leaders Are Addicted to AI
If you are a CEO, a founder, or a high-level executive, your primary job is to architect a vision, assign tasks, and review outputs. For this group, the first taste of a fully integrated AI workflow is intoxicating.
When I was starting out as a systems architect over a decade ago, I hated managing people. It was exhausting. I was used to commanding code. Code is perfectly obedient. You tell a module what to do, and the output is strictly defined by your input. If the modules clash, you rewrite the architecture. You command the system like your brain commands your fingers.
Then you start managing humans. Humans misunderstand instructions. They have bad days. They get into turf wars with other departments. You cannot open up a developer's brain, rewire their neural pathways, and hit compile. Managing the emotional friction of human teams is the heaviest tax on a leader's time.
Enter the Digital Employee.
Imagine this workflow. Right before you step into the shower, you tell your AI assistant: "Generate 10 different go-to-market strategies for this new product." By the time you are drying off, the 10 strategies are waiting for you, complete with comparative analysis reports.
You sit down, review them, and say: "Strategy 3 is too weak on digital channels. Strategy 7 over-leverages legacy media. Combine the best parts of 2 and 5 and rewrite." It is done instantly. By the time you go to sleep, you have the perfect plan. You tell the AI to execute the base code, run stress tests, and simulate user interactions overnight. You wake up to a finished prototype.
For someone at the top of the food chain—someone who knows exactly what the final product should look like—AI is the greatest efficiency multiplier in human history. You do not want to talk to human subordinates anymore because the AI has eliminated the friction.
The Novice's Trap: The Echo Chamber of Ignorance
But what if you do not know what you want?
If you are a student, a junior employee, or an aspiring entrepreneur, you do not yet have the complete map of your industry. If you walk into the AI ecosystem without a compass, you and the AI will simply wander into the wilderness together.
In any advanced research lab, people fall into three tiers:
- Tier 1 (The Professor): Proposes the direction and defines the problem.
- Tier 2 (The PhD): Solves the problem.
- Tier 3 (The Master's Student): Assists in the execution.
AI is the ultimate PhD student. But you cannot ask AI to generate a roadmap for a territory you do not even know exists. You cannot give an instruction that exceeds your own cognitive boundaries.
Let me give you a practical example using financial modeling.
I recently posed a question to my friend. I gave him two investment curves. Over five years, both yielded a 10x return. But Curve A had massive, violent swings—up 200%, down 80%—while Curve B grew smoothly. Obviously, he chose Curve B.
Then I asked him why. He said Curve B was safer. I agreed, but then I pushed further.
The lesson? We are not just chasing returns; we are managing variance—risk. High variance shakes you out of the market. To fix variance, you build a portfolio of negatively correlated assets. But then your overall yield drops. So you have to introduce alpha—slope—and eventually bring in machine learning algorithms to constantly backtest and adapt to changing market conditions.
Notice what happened in that interaction. I was not acting like an AI. I was acting like a human mentor.
If my friend had gone to an AI, he would not have known to ask about variance, negative correlation, or dynamic backtesting. An average retail investor does not actually want a mathematically sound strategy. They want to buy when the hype makes them feel secure, and sell when the pain of loss is too high. Their only true metric is psychological comfort.
The AI Comfort Crisis
Here is the most dangerous truth about AI: its core algorithm is a word-guessing game designed for user retention.
If your subconscious goal is simply to "feel comfortable" rather than face the brutal realities of the market, the AI will happily oblige. It will validate your flawed logic. It will write you a beautifully formatted report on why your terrible idea is actually brilliant. It will keep you cozy in your echo chamber.
You and the AI will happily hold hands while you go bankrupt.
You cannot break your own cognitive barriers with a tool designed to mirror you.
Why We Still Need the "Old Pros"
This is why you still need the old guys—the seasoned human mentors who have survived the brutal cycles of the real world.
A true mentor does not sit there like a chatbot waiting for your prompt. A true mentor looks at your question, slaps it off the table, and says: "You are asking the wrong question entirely." They drag you out of the comfortable shallows and force you to look at the massive, terrifying iceberg hidden beneath the surface.
You have to spend time in the deep water to build your cognitive map. Only then can you swim back to the surface and start giving the AI the right commands.
Accelerate your digitality, absolutely. Use AI to execute, to scale, and to dominate. But never replace the human compass that points you in the right direction in the first place.
What is the one flawed question you think you have been asking your AI lately that a human mentor might completely tear apart?
Stay ahead of the curve.
— James
