Steve Jobs killed the stylus by giving us back our fingers. Twenty years later, AI is doing the same thing to business data. The interface is reverting to plain English, and that changes everything about who can ask questions and what they can know.
I remember the moment vividly. January 2007, Steve Jobs on stage introducing the iPhone. The obvious question hanging in the room: how do you interact with a touchscreen? Every smartphone at the time used a stylus and a plastic keyboard. Jobs posed the question himself, then answered it: "Who wants a stylus? You have to get 'em, put 'em away, you lose 'em. Yuck. Nobody wants a stylus." And then the pivot that changed everything. We are going to use the best pointing device in the world. One we are all born with. We have ten of them. (My paraphrase. The spirit is exact.)
The insight was not about eliminating a plastic stick. It was a statement about interface design: the best interface is the one that requires no learning, because the human already has the skill. Touch was always how people interact with the physical world. The technology just needed to catch up.
A different story, same principle. Bill Belichick, arguably the most analytically rigorous coach in the history of professional football, famously threw his Microsoft Surface tablet on the sideline bench and went back to printed black-and-white photographs. His complaint, paraphrased, was that there was simply too much between him and what he needed to see. A device, a cloud connection, a camera feed, a software interface, screen glare on a bright afternoon. Each layer adding friction between intention and insight. The old printout was still faster.
Easy to misread that as a technophobe's resistance. It is actually a precision critique. When there is more technology between the human and the answer than the answer is worth, the technology loses. The interface was the problem, not the information.
The best tool is the one that disappears. The moment you are thinking about the tool, it has already failed you.
Those two stories are playing out again right now, in business data. For two decades, getting insight from enterprise systems required learning a machine's language first. SQL. DAX. Tableau. Power BI. Each tool added a layer between the person with the question and the system with the answer. Most people stopped asking. The stylus won, and curiosity lost.
AI is the touchscreen moment for data. You can now ask a question in the same language you learned before you could read. Misspelled, conversational, imprecise. It does not matter. The barrier to information is gravitating back to the limitation of the question itself, not the limitation of the tool. That is a completely different problem. A more human one. And a far more exciting one.
Kahneman would add a footnote. In Thinking, Fast and Slow, he distinguishes System 1 (fast, intuitive, automatic) from System 2 (slow, deliberate, effortful). Business analytics has been a forced System 2 exercise for twenty years. The tool friction was exhausting, but it was occasionally protective. By the time the analyst came back with the report, you had sometimes reconsidered the question. AI removes that pause. The CFO who asks "why is our margin softening?" gets a confident answer in seconds. Whether or not the premise was sound. System 1 is also where confirmation bias lives. The limitation is now the quality of the question, not the capability of the tool. That is more interesting than it is alarming. Worth knowing as you hand the interface back to the humans.
There is a lot of technology between the user and the answer in an AI-powered data platform. More than ever, actually. But like Jobs' iPhone, it has become invisible. The user just asks. We are genuinely excited to be a small part of this industrial and human revolution. The stylus is gone. Nobody is going to miss it.
Natural language querying is only as good as the data it runs on. Marquis IQ connects every ERP, normalizes master data, and builds the clean layer that makes AI answers trustworthy.