When AI asks a better question than your boss.
The smartest use of this technology I have seen was not a faster answer. It was a question the operator had not thought to ask — and the reason most teams never get one.
It was not the tools. It was the questions.
I was on a call with a startup leader navigating the messy middle of AI adoption. Three lines of business, real potential, and a team split down the middle: marketers embracing the tools, engineers dragging their feet, leadership stuck somewhere between.
What struck me was not the stack they had assembled. It was the questions they were not asking.
A children's book, and the question underneath it.
At one point he told me about a kids' book he had co-written with AI — a big vehicle that wants to be a race car, or an airplane, until it learns to embrace what it was built for. He fed the model writing samples, gave it creative direction, iterated on the draft. Then the machine asked him this.
That is the whole point. This is not primarily about saving time. It is about surfacing better questions, faster than we would reach them on our own.
We have been burned before, and it is a liability.
When people talk about AI now they are mostly either panicking or posturing. I have seen this movie. I remember Napster — everyone loved it, then it vanished, and two decades later the name quietly sold for $207 million. Value gets created and destroyed long before most companies understand what they are looking at.
Our generation has been burned by cycles before: crypto, NFTs, the metaverse. Each time we moved too fast, too slow, or missed it. That fatigue is real. It is also a liability, because it is now the main reason capable operators sit out a shift that is not a cycle.
The best use I have seen was in a transcript.
There are tools now that analyze meetings for tone, engagement, and sentiment — how inclusive your language was, how engaged your team was, where the time went. I looked at one client's readout and said: you realize you are late to every meeting, and spend a third of each one describing problems you caused?
They had nothing to say. The mirror does not lie.
What companies ask for
What changes anything
Your data will not contradict you unless you ask it to.
That same team had just spent a large sum on external research, and the conclusions landed flat — in the founder's words, “a regurgitation of what we already knew six years ago.”
Here is the shift. Your data will not give you a new idea unless you ask it to contradict you. You need systems that flag what you did not think about, rather than systems that summarize what you did.
“Make me think about what I'm not thinking about.”
That is the prompt. That is where the edge is. Everything else is automating the same mediocrity at speed — and the brands that win have never been the ones with better features, they are the ones that make someone feel something.
It does not replace you. It reveals you.
Most people think this technology is about replacing people. It is not. It is about revealing where you are not as sharp as you thought. Where your team is stuck. Where your questions are too shallow. Where your culture is too safe to be corrected.
The opportunity is not in adopting tools. It is in building a practice where the machine is allowed to ask the questions nobody in the room is willing to raise.
Build the practice, not another dashboard.
The operators who get a real answer out of this technology are the ones who set it up to argue with them. That structure is buildable, and it is most of what we do.