Let’s be brutally honest. Most companies are getting AI wrong.
We are seeing the friction of trying to fit a 21st-century sense into a 20th-century business model.
They’re spending millions on a new technology and getting… faster spreadsheets? More polite chatbots? Marginally better marketing plans?
The initial hype is clearly dead, and for many leaders, the ROI is looking dangerously thin.
Why?
Because we’re asking it stupid questions.
We are treating the most powerful perceptual tool ever invented like a slightly-smarter, cheaper intern. We ask it to summarize reports, fetch data, and write boilerplate emails. We are focused on optimization.
Not only is this incredibly inefficient, but it is a (borderline embarrassing) failure of imagination.
Granted, this isn’t really our fault. Our entire mental model for business was forged in the Industrial Revolution. We think in terms of efficiency and inputs/outputs. We hire “hands” to do work, and we look for tools to make that work faster.
So when AI showed up, we put it in the only box we had: the faster intern box.
We ask it Industrial-Age questions: “How can you write my emails faster?” “How can you summarize this 50-page report?” “How can you optimize my marketing budget?“
These aren’t necessarily bad questions. They’re just lazy. They are requests for efficiency.
They are limited, and they are all, at their core, asking how to do the same things, just faster. As long as we are only optimizing, we are missing the potential revolution entirely.
The real breakthrough, I believe, will come when we stop seeing AI as an extension of our hands and start seeing it as an extension of our perception.
Your intern is an extension of your existing ability. They do the tasks you assign, just saving you time. A new sense, like infrared vision, grants you an entirely new ability (thinking X-Men). It lets you see heat, a layer of reality that was completely invisible to you before.
This is the shift we’re missing.
Its true power isn’t in doing tasks we already understand, but in perceiving the patterns, connections, and signals we are biologically incapable of seeing. Humans are brilliant, but we are also finite. We can’t track the interplay of a thousand variables in real-time. We can’t read the unspoken sentiment in ten million data points.
AI can.
When you reframe AI as a sense, the questions you ask of it change completely. You stop asking about efficiency and you start asking about insight. You stop asking, “How can I do this faster?” and you start asking, “What can I now see that I couldn’t see yesterday?” For example, perceiving a hidden market anxiety.
So what does this shift from intern to “sense” actually look like?
It comes down to the questions we are asking. The quality of our questions is the limiting factor, not the technology (for the most part).
Look at marketing. The old paradigm, the “intern question,” is focused on automating grunt work:
The Intern Question: “AI, write 10 social media posts and five blog titles for our new product launch.”
This is a request for efficiency. It gets you to the same destination a little faster.
But the organ question is a request for perception:
The Organ Question: “AI, analyze the 5,000 most recent customer support tickets, forum posts, and negative reviews for our competitor. What is the single unspoken anxiety or unmet need that connects them?”
The first answer gives you content. The second answer gives you your next million-dollar product.
Let’s look at strategy. The intern question is about summarization:
The Intern Question: “AI, summarize the top five industry trend reports for the next quarter and give me the bullet points.”
This is a request for compression. You’re asking AI to act as an assistant, saving you a few hours of reading.
But the “Organ Question” is about synthesis and signal detection:
The Organ Question: “AI, find the hidden correlations between global shipping logistics, regional political sentiment, and our own internal production data. What is the emergent, second-order risk to our supply chain in six months that no human analyst has spotted?”
The first question helps you prepare for the meeting. The second question helps you prepare for the future.
To summarize all of this word vomit, I believe that the AI revolution isn’t stalling; more so waiting. It’s waiting for us to catch up.
What we are seeing is the friction of trying to fit a 21st-century sense into a 20th-century business model.
We are all still pointing this new tech at our old problems; our inboxes, our reports, our slide decks. We are asking it to help us optimize the past, when its real ability is to help us perceive the future.
The most critical skill for leadership in this new era will not be prompt engineering. It will be question design. Stop asking “How can AI do this job faster?” and start asking, “What new work becomes possible because we can now see in this new way?”
So, ask yourself and your team: Are you using AI to get better answers to your old questions? Or are you using it to find entirely new questions?









