Why Some People Get So Much More Out of AI Than Others
Two people can use the same AI tool and come away with completely different impressions.
One of them says it has changed how they work, helping them draft reports faster and freeing up hours every week. The other tried it a couple of times, got some mediocre answers, and decided it was overhyped.
What's strange is that intelligence doesn't seem to predict which camp someone lands in, and neither does technical skill. I've seen very capable professionals struggle with AI, and people with no technical background use it well.
The thing that usually separates them is expectations.
Most People Expect Either Magic or Disaster
One group expects AI to solve hard problems instantly, produce flawless work, and take over big chunks of what they do. The other expects it to gut their expertise and make human judgment beside the point.
Both are off the mark, and both tend to end in disappointment.
The gap between what people expect and what they get shows up in the research. A 2026 National Bureau of Economic Research study surveyed nearly 6,000 senior executives in the U.S., U.K., Germany, and Australia. Despite most of them using AI, 89% said it had made no measurable difference to their productivity (Yotzov et al., 2026). The tool wasn't really the problem; their expectations were.
It's a little like handing someone a calculator and watching them expect it to think for them. A calculator doesn't think. It runs calculations accurately within a narrow range, and once you understand that, it's useful right away. Expect something else from it and it just looks limited.
We're working through the same confusion with AI now, though the stakes are higher.
Better AI Makes Judgment More Important, Not Less
This part tends to surprise people: as AI gets more capable, your own judgment matters more, not less.
If AI writes a first draft, you still have to decide whether it's right. If it summarizes a report, you have to check that it caught what mattered. And when it suggests a course of action, you're the one who has to work out whether that holds up in the real world.
The skill was never about accepting what the AI gives you. It's about knowing how to judge it.
There's good evidence for this. Economists at Stanford and MIT, writing in the Quarterly Journal of Economics, followed more than 5,000 customer support agents who were given access to an AI assistant. Productivity went up 15% on average, but the benefit wasn't spread evenly. Newer workers improved a lot, while the most experienced ones barely changed. The AI mostly helped people catch up to a level the experts had already reached on their own (Brynjolfsson et al., 2025).
In other words, it speeds up the climb toward expertise without standing in for it.
The Best Use Isn't the One Most People Try First
People usually reach for AI to help with writing first: drafting an email, summarizing a document, putting together a report. Those uses are real, but they aren't where the biggest payoff tends to be.
The people who get the most out of AI treat it as a thinking partner. Rather than asking it to produce something, they ask it to challenge something:
What am I missing?
What am I assuming without realizing it?
What's the best argument against my position?
What would someone who knows this field take issue with?
Microsoft's 2025 New Future of Work Report, which pulls together findings from a wide range of studies, reported that AI framed this way, as something that builds on human thinking rather than replacing it, tends to get adopted more readily and produce better results than AI sold purely as a way to save time (Butler et al., 2025).
The two approaches pull in different directions. One hands your thinking off to the machine. The other puts your thinking under pressure and tends to improve it.
Know What It's Good At, and What It Isn't
Large language models are good at explaining ideas, drafting text, brainstorming, reframing a problem, and putting complicated things into plain language.
They're a lot less dependable when it comes to checking facts, doing exact calculations, giving legal or medical advice, or telling you whether something is true rather than merely plausible-sounding.
Stay inside what it's good at and you'll probably be impressed. Push it for certainty when it's built to deal in probabilities and you'll probably be let down. Telling those situations apart isn't really a technical skill. It's mostly a matter of calibration.
The Skill Worth Building
The specific tools will keep changing, and so will the interfaces and the model names. What won't change is the underlying ability to work well with these systems.
That means asking sharper questions, giving useful context, reading the output with a critical eye, and noticing when the tool is helping you rather than quietly steering you wrong.
The research here is encouraging. A study in Information Systems Research found that when people moved to a more advanced AI model, only about half of the improvement came from the model itself. The rest came from how users changed the way they asked their questions. And the people who did this best weren't software engineers; they came from all sorts of backgrounds and ages. As the researchers put it, prompting well has more to do with communication than coding (Jahani et al., 2026).
Those habits carry over from one generation of tools to the next.
The more useful question, then, isn't whether AI will replace people. It's what happens when people who know how to use it start working alongside people who don't.
That's already playing out in workplaces every day. The ones pulling ahead aren't necessarily the most technical. They just have a realistic sense of what the tool can and can't do, plus enough practice to tell the two apart. None of that requires a computer science degree or learning to code. It takes a clear picture of the technology and some time spent using it, which is a smaller ask than most people assume.
If you are ready to get that clear picture and start leveraging AI the right way, that is exactly why I wrote Finally, AI Makes Sense.
Inside, I break down the practical frameworks, communication habits, and real-world strategies you need to turn AI into your most powerful asset—written specifically for normal, busy professionals, not software engineers.
References
Brynjolfsson, E., Li, D., & Raymond, L. (2025). Generative AI at work. The Quarterly Journal of Economics, 140(2), 889–942. https://doi.org/10.1093/qje/qjae044
Butler, J., Jaffe, S., Janßen, R., Baym, N., Hecht, B., Hofman, J., Rintel, S., Sarrafzadeh, B., Sellen, A., Vorvoreanu, M., & Teevan, J. (Eds.). (2025). Microsoft new future of work report 2025 (Microsoft Research Tech Report MSR-TR-2025-58). Microsoft Research. https://aka.ms/nfw2025
Jahani, E., Manning, B. S., Zhang, J., TuYe, H., Alsobay, M., Nicolaides, C., Suri, S., & Holtz, D. (2026). Prompt adaptation as a dynamic complement in generative AI systems. Information Systems Research. https://doi.org/10.1287/isre.2025.2029
Yotzov, I., Barrero, J. M., Bloom, N., Bunn, P., Davis, S. J., Foster, K. M., Jalca, A., Meyer, B. H., Mizen, P., Navarrete, M. A., Smietanka, P., Thwaites, G., & Wang, B. Z. (2026). Firm data
Welcome to Azimuth Line Press: Mapping the Stories Ahead
May 17
Welcome to the very first post on the Azimuth Line Press blog! Whether you stumbled upon us by chance or have been following our journey toward launching this publishing imprint, we are incredibly grateful you are here.
An "azimuth line" is all about direction, navigation, and finding a precise path forward. That is exactly what we aim to do with this space. We aren't just here to announce new book releases; we want to build a community of readers and thinkers who love diving deep into the written word.
What to Expect on This Blog
We know your inbox and your reading feeds are crowded, so our promise to you is simple: every post here will aim to add real value to your day. Moving forward, the core of this blog will feature exclusive book excerpts paired with behind-the-scenes insights. Instead of just dropping a chapter and walking away, we will be sharing:
The "Why" Behind the Words: Context on what inspired specific topics, methods, or arguments.
Deep Dives: Thoughtful commentary that expands on the themes in our books, connecting them to broader ideas, history, or everyday life.
Author Perspectives: Glimpses into the creative process and the research that goes into bringing our titles to life.
Join Us for the Journey
Think of this blog as an open conversation. We want to hear your thoughts on the excerpts we share, your questions about the writing process, and your own reflections on the themes we explore.
To make sure you never miss an excerpt or an update, feel free to bookmark this page or sign up for our newsletter below.
Thank you for joining us at the starting line. We can't wait to share our first official excerpt with you very soon. Until then, happy reading!