Work, Wonder, and the Machine

Introduction

An Invitation

A new tool can make an ordinary workday feel newly uncertain. You may wonder whether AI could help with a task you find tedious, whether you are expected to use it, or whether a confident answer is one you can trust. Those are not signs that you are behind. They are sensible questions to bring to a technology that can produce useful work and still make mistakes. You do not need to be a programmer, an enthusiast, or a person who has already chosen a side. Curiosity is enough to begin.

There is a practical reason to pay attention: for some people, AI is already part of how work gets done, while many others use it little or not at all. In a Pew Research Center survey, 16% of workers said at least some of their work was done with AI; most workers said they used AI little or not at all at work.^1,2That gap is not a verdict on anyone’s ability. It is an invitation to understand what the tools can do, what they cannot, and where your own experience and judgment matter.

The benefits people report are real, but they are not all the same. Among workers who had used AI chatbots for work, 40% said the tools were extremely or very helpful for doing things more quickly, while 29% said they were that helpful for improving work quality.³Speed is useful, but it is not the same as accuracy, care, or a good decision. This book takes both sides seriously: a tool may help you move through a draft or find information, while the final work still deserves a human review.

It also makes sense to feel uneasy. In a Pew survey, 52% of U.S. workers said they were worried about AI’s future impact in the workplace.¹A worrying headline can make the future feel settled, but concern is not a forecast, and a possibility is not a certainty. You deserve more than either easy reassurance or alarm. You deserve a way to examine changes in your own work, ask better questions, and make choices based on what you can observe.

That is the spirit of Work, Wonder, and the Machine. I’m Shad Eddie Mongrue, and I’ve written this as a practical guide, not a sales pitch for technology. You will learn to choose tasks thoughtfully, give clearer instructions, build repeatable workflows, and keep your judgment in the loop. The aim is not to hand your work over to a machine. It is to discover whether a tool can help with a particular part of the work while leaving room for what only you can contribute.

This subject is worth loving—not because machines solve everything, but because learning to work with them can make room for better questions, more curiosity, and sometimes less tedious effort. A useful experiment might save you time; another might teach you that a task needs your attention after all. Both are worthwhile discoveries. You can begin with one ordinary piece of work, stay skeptical without becoming closed off, and let practice—not pressure—shape what you think.

Reasons to fall in love with this subject

• Curiosity gives you a way in You do not have to arrive with technical talent or a perfect prompt. You can start by noticing a small snag in your work and asking whether AI might help with one part of it. A useful question can be simple: What would make this draft clearer? What information am I missing? Curiosity turns experimentation into something you can practice, not an exam you pass or fail.
• Small tasks make change easier to see A job can feel too large to evaluate, but a task is something you can examine closely. You can consider whether a tool might help organize notes, summarize information, or create a first draft, then decide what still calls for your knowledge and attention. This way of looking at work helps you avoid sweeping conclusions about whole occupations and keeps your choices grounded in the work in front of you.
• A draft can be a beginning, not a verdict Generative AI can produce fluent text, but fluency alone does not establish that an answer is correct or suitable. That distinction is useful: you can treat an output as material to inspect, question, revise, or set aside. Instead of expecting a magic answer, you can ask for a starting point and then bring in your subject knowledge, the facts you can verify, and the context the tool does not know.
• Better workflows make the human role visible The interesting question is not simply whether AI can do something. It is how the tool fits into a process: what you give it, what you ask it to produce, and where you pause to check the result. A documented study of customer-support agents found an average productivity increase of 14% after access to an AI tool, with a 34% improvement for novice and low-skilled workers and minimal impact for experienced and highly skilled workers.⁴That result is informative, not a promise for your job.
• Your skills still have room to grow Learning to use a tool does not require you to stop developing communication, critical thinking, collaboration, or subject knowledge. In fact, those abilities help you judge whether an output makes sense and explain what needs to change. Pew found that about 70% or more of workers considered interpersonal skills, communication, critical thinking, and basic computer skills very or extremely important, regardless of their AI use.²You can practice with AI while strengthening the skills that help you use it well.
• Thoughtful practice can make uncertainty manageable You cannot control every change in the workplace, but you can learn to notice how a tool affects your own tasks, what rules apply, and what you would like to learn next. A little structured practice can replace vague pressure with a concrete question and a next step. You may discover a useful shortcut, a reason to keep a task human-led, or a limitation that changes your plan. Each outcome helps you make a more informed choice.

How to use this book

You can read the chapters in order, since they move from understanding workplace change to choosing tasks, giving instructions, and reviewing results. Or begin with the chapter that answers the question on your mind. You do not need to master every idea before trying one small, low-stakes task. Keep your own work context in view: a tool that helps with one kind of task may be a poor fit for another.

When you practice, write down what you want the tool to help with, what information it needs, and what a good result would look like. Try a clear request, inspect the response, and revise your instructions or your approach. Notice what improved and what still needed your attention. If the result is not useful, that is information—not proof that you asked badly or that you must keep trying.

Before using AI for work, check the rules and expectations that apply where you work. Do not enter confidential information into a tool unless it has been approved for that use. Keep a person responsible for decisions that require judgment, context, or verification. The goal is not to use AI everywhere; it is to make deliberate choices about when it helps and when it does not.

Chapter 1

A New Kind of Workday

The workday is changing, but not all at once

AI matters at work because it has entered the conversation about ordinary tasks: drafting, researching, summarizing, organizing information, and preparing for decisions. It can feel as though a new kind of workday has arrived everywhere at the same time. The evidence points to a more uneven picture. Most American workers say they use AI little or not at all in their jobs; 16% say at least some of their work is done with AI, including 2% who say all or most of it is². Another 25% say some of their work could be done with AI, though they are not using it much now⁵. Possibility and everyday practice are not the same thing.

That gap matters. You may hear about a tool that seems capable of a great deal and still have no approved way to use it at your workplace. Or you may already use a tool for one small task without thinking of yourself as someone who “uses AI.” Your experience is not a referendum on whether the technology is important. It is one point in a transition that differs across jobs, employers, and tasks.

Uncertainty is part of the story, too. In an October 2024 survey, 52% of U.S. workers said they were worried about AI’s future impact in the workplace¹. That worry is understandable. A tool can be useful and unsettling at once. It may save effort on one task while raising questions about who reviews the result, how work is valued, or what skills will matter next. You do not have to settle those big questions before you can learn something useful about your own work.

A headline is a question, not a forecast

Headlines often compress a complicated change into a clean dramatic claim: jobs are disappearing, work is being transformed, or a new tool will make everything faster. Treat those lines as invitations to investigate, not as reliable predictions about your own next year. A general trend does not tell you what will happen to a particular role, team, or task.

Even employers’ expectations about skills are not a timetable for your personal working life. The World Economic Forum reports that employers expect 39% of key skills required in the job market to change by 2030⁶. That is a reason to stay curious and keep learning. It is not proof that 39% of jobs will vanish, nor a list of exactly what you must learn tomorrow. A skill can change in importance while the work itself is reorganized, supported by new tools, or combined with other responsibilities.

A helpful question is more specific than “Will AI take my job?” Try: “Which parts of this job might change, and what would that mean for the parts that still need human judgment?” Break the role into activities. A person may spend time locating information, preparing a draft, checking details, explaining a recommendation, and responding to another person’s needs. AI may be relevant to some of those activities and a poor fit for others. The answer depends on the task, the quality of the tool, the stakes, and the rules around its use.

This approach does not promise that every worker or workplace will benefit. It simply gives you a more useful scale for thinking. A role is not one indivisible block. It is work made up of parts, and those parts can change at different speeds.

Start with friction, not fascination

When a new tool appears, it is easy to begin by asking what it can do. For your own work, begin instead with a recurring point of friction. Which task makes you say, “I wish I had a first draft,” or “I spend too long finding that information,” or “I keep reorganizing these notes before I can think”? Naming the friction gives you a practical experiment. It also keeps you from handing the tool a task simply because the task looks impressive when automated.

Suppose you regularly turn a pile of notes into a short update. You might ask an approved AI tool to propose an outline or identify possible themes. Then you can check whether it missed an important qualification, distorted the meaning, or made an unsupported claim. The goal is not to make the tool responsible for the update. The goal is to see whether it can help with one bounded part while leaving you in charge of what the update should say.

Workers who have used AI chatbots for work report more help with speed than with quality: 40% say the tools have been extremely or very helpful in doing things more quickly, while 29% say they have been highly helpful in improving work quality³. Those figures do not tell you what will happen in your job. They do suggest a useful distinction for your own trial: Did the tool reduce time, improve the result, do both, or do neither?

Keep the first test modest. Choose something repeatable and low-stakes, define what a good result would look like, and decide what you will check yourself. If the task involves sensitive information, follow your workplace rules and do not enter confidential material into a tool that has not been approved. A small, careful test can teach you more than an ambitious experiment with no clear way to judge the outcome.

Use uncertainty to ask better questions

Curiosity is not a technical credential. It is the habit of noticing what you do not yet know and asking a question that can help you find out. You can be curious about the tool, but you can also be curious about your own work: What part is repetitive? What part depends on context? Which errors would matter? What would make a response genuinely useful to the person receiving it?

Those questions turn vague excitement into a workable plan. Instead of asking an AI system to “make this better,” you might ask it to offer three possible structures for a short update, using only the information you provide. You can compare the structures, reject one that changes the emphasis, and revise the one that fits. The experiment is not a test of whether you are clever enough to prompt a machine. It is a way to learn what kind of help is useful in a particular situation.

It is also a way to protect space for the parts of work you value. If a tool can help you get past a blank page or sort a long set of notes, you might have more attention for deciding what matters, asking a colleague a thoughtful question, or explaining an idea clearly. That is a possibility, not a guaranteed result. Sometimes the tool creates extra checking or produces a draft that is more trouble than it is worth. Noticing that is part of learning, too.

You do not need to love every tool, or use one for every task, to find value in this subject. The promising part is the chance to become more deliberate: to spot tedious work, test a practical alternative, and keep the decisions that matter in human hands. When a headline makes the future sound settled, curiosity gives you permission to look closer.

A life example

Myths and honest truths

• Myth: If a headline says AI can do a job, that job is about to disappear. Truth: A headline about a capability does not predict what will happen to a whole occupation. Ask which tasks are involved, what the evidence actually measures, and how people will use the tool.
• Myth: If AI saves time, it must improve the work. Truth: Speed and quality are different outcomes. A faster draft can still need careful checking, revision, or rejection.
• Myth: You need to be tech-savvy to experiment usefully. Truth: A clear question about a familiar task is a good beginning. Curiosity and sound judgment matter more than performing technical wizardry.