You have Einstein right here, but you're making him your secretary!

Sometimes I feel like I'm "carrying Einstein around with me"; I'm aware of the intelligence of artificial intelligence. But the painful truth is: if I use that intelligence like a secretary, my gains are limited. What changes isn't the level of intelligence; it's where it's used .

Today, many people use artificial intelligence for tasks such as: “Fix my email.” “Improve the presentation.” “Give me 10 ideas.” “Make a 3-day trip to this city.”

Are these bad? No. In fact, they're great for starting out. But in 2025, this usage will be like saying "I know Excel" in the workplace: necessary, but it won't set you apart on its own. Because most of it is packaging: speeding up the packaging process—more polite sentence, cleaner slides, more fluid text.

Think of it this way: You have a Swiss Army knife, but you're only using the nail clippers. That gets the job done, but that's not what a knife is for.

The real difference begins when you take artificial intelligence from the "production line" and put it on the "decision line".

Production line: print, edit, summarize, design. Decision line: set criteria, generate scenarios, identify risks, test assumptions, suggest measurements, ask "what would change my mind?".

Let's take the email example. Most of us say "to refine" it. This is like polishing a plate: it shines. But in high-level communication, the real question is: What will this email change? Will it lead to a decision? Will it reduce a risk? Will it resolve an uncertainty?

Without a clear "I want this decision," a clear "there's this risk," a clear "if you approve this, we'll release it on this date," even the best-written email in the world will remain just a well-written email.

Presentations are more ruthless. Most presentations look nice, but they lack a single clear decision. Artificial intelligence enhances slides like boosting the lights in a shop window: the window shines; but if the product being sold isn't clear, nobody will go to the cashier. What makes a presentation good isn't transition animations; it's three sharp messages and a clear explanation of "if we do this, this metric will change."

The travel plan metaphor fits perfectly here. Getting a "3-day Rome" plan is sweet, but it's usually an average prescription that everyone can eat. The real value lies in the doctor's questions: "What are you allergic to? How much can you walk? What's your energy level? Is your goal to relax or explore?" If you don't give artificial intelligence context, it will give you an average route; in business, if you don't give it context, it will produce an average output. And an average output... an average impact.

A concrete example: The project is delayed. If you say "Write me an email about the status," I'll write one.

But the real leverage is this: “What are the 5 root causes of the delay? Which is most likely? Which will have the greatest impact? What are the 3 interventions that will have the greatest impact with the least effort? How can a manager figure this out in 30 seconds? On which metric will we see success?”

This isn't about writing emails; it's about establishing the management behind the email. What transforms AI from a secretary to an analyst isn't the output, but the set of questions you ask.

It's the same with article writing. When you say, "Write an article on this topic," a text comes out. But a good text comes from these questions: "What is the claim of this article? Where will the reader object? What counter-arguments will be made? Which examples will make them say 'okay'? Where will they get bored, where will they be convinced?" Here, artificial intelligence is not a text generator, but a thought testing device—it works like a mirror; it shows the loose screws in your argument.

And the most critical point that nobody talks about: it's not about getting one good result from AI, but about making it a routine . Like creating a home layout instead of tidying up the same mess every week. Weekly executive summary, risk radar, decision note, watchlist… Once a system is established that works, an invisible acceleration begins within the team: meetings get shorter, uncertainty decreases, and surprises diminish.

This is exactly what we mean by "repositioning yourself": moving from using AI as an "accelerator" to using it as a leverage. Because in a few months, everyone will be getting their emails corrected, everyone will be getting their presentations improved. There will be no advantage left. The advantage will remain in the quality of your thinking: setting the right criteria, providing the right context, choosing the right measurement, and building the system.

The bottom line is clear: If you feel like you have Einstein by your side, don't send him for photocopying. Get him to sit at the decision-making table. Have him ask questions like, "Why is this option good?", "Where does this risk come from?", "What would change my mind?", "How will I measure success?"

If you could only make one move today: Choose the one task that consumes the most of your time (reporting, project tracking, customer communication, meeting cycle). Ask AI for not just output, but also criteria, risk assessment, measurement, and routine analysis.

Question:

Are you using AI primarily for "packaging," or have you placed it in the "decision-making process"?

Write your role and your most time-consuming task in the comments; I'll leave you with a prompt that transforms you from "secretary to analyst".