Step 5 of 6 · Build Confidence In Uncertain Times
Working Alongside AI
Working Alongside AI
Step 5 · 12 min
🎬 Video lesson coming soon
We've talked about what AI cannot do. We've talked about identity and growth. Now I want to get practical.
Because alongside the genuine anxiety about AI, there is also an opportunity — one that is already available to you, regardless of your field, your background, or how technically minded you are.
The opportunity is this: to become someone who works skillfully alongside AI, rather than against it or in fear of it.
Not a programmer. Not an AI expert. But someone who understands how to direct these tools effectively, evaluate their outputs with human judgment, and add the thing they can't — your experience, your values, your understanding of the actual human situation.
The professionals who will thrive in the next ten years are not exclusively the ones who built the AI. They are the ones who use it well — who know when to trust it and when not to, what to ask of it and how to improve on what it gives them.
That is a learnable skill. And it doesn't require you to become a different person.
Augmentation, Not Substitution: MIT economist David Autor's research on technology and labour shows that automation historically augments human labour in high-skill tasks rather than substituting it — amplifying what skilled humans can produce rather than replacing them. The most economically valuable position in an AI-augmented world is not competing with AI but directing, evaluating, and adding genuine human judgment to AI-generated outputs.
The Human-in-the-Loop Advantage: AI systems in high-stakes contexts — medicine, law, engineering, education, governance — are most effective and safest when humans remain actively in the loop: providing context, evaluating outputs, catching errors, applying ethical judgment, and taking responsibility for decisions. This 'human-in-the-loop' position is not going away — it is increasingly valuable as AI capabilities expand and the stakes of AI errors rise.
AI Literacy as a Multiplier: Research on AI adoption in professional contexts shows that professionals who understand AI capabilities and limitations — not at an engineering level, but at a user-literacy level — consistently outperform peers who either avoid or uncritically accept AI tools. AI literacy is not about coding; it is about knowing when AI is reliable and when it isn't, how to direct it effectively, and how to add the human judgment it lacks.
MIT economist David Autor has spent years studying how technology affects labour markets. One of his key findings disrupts the simple 'AI replaces workers' narrative.
Technology, historically, has been most powerful when it augments skilled human work rather than substituting it. What tends to happen is that automation handles the routine, rules-based work within a field — and the human professionals who remain then work with the technology to produce outputs that neither could produce alone. The surgeon with robotic assistance. The analyst with data modelling tools. The teacher with adaptive learning platforms.
The economically valuable position in this landscape is not the person doing the manual work that AI can do. It is the person who can direct the AI, evaluate its outputs, and apply the human judgment that determines what to do with them.
In high-stakes domains — medicine, law, education, financial advice, engineering — there is a concept called 'human in the loop.' The idea that, for decisions with significant consequences, a trained human must remain actively involved: reviewing, approving, catching errors, applying contextual judgment that the AI cannot.
This position — human in the loop — is not a vestige of the pre-AI world that will eventually be automated away. It is increasingly valuable as AI capabilities expand, precisely because the consequences of AI errors also expand. The more powerful the tool, the more important the skilled human who directs and evaluates it.
What this means practically is that AI literacy — not engineering knowledge, but user literacy — is one of the highest-leverage things a professional can develop right now. Understanding what AI tools are good at and what they aren't. Knowing how to prompt them effectively. Knowing how to evaluate whether their output is accurate, appropriate, and ethically sound. Knowing when to trust them and when to override them.
This is not technical wizardry. It is the same kind of judgment you apply to any tool or any junior colleague. You understand what they're good at, you assign work accordingly, and you maintain the oversight and final judgment that your experience provides.
You already know how to do this. The tool is new. The skill is not.
Find a comfortable position · Read slowly
Two things.
First: a curiosity experiment.
In the next three days, use an AI tool — any one you have access to — for a task in your actual work or life. Not to replace your judgment, but to assist it. Ask it a question relevant to something you're working on. Use it to draft something you'd then edit. Ask it to summarise something complex and then evaluate whether it got it right.
Then write three observations: — Where was it useful? — Where was it wrong, incomplete, or missing something important? — What did your judgment add that the AI couldn't?
That third question is the important one. Because what you add — the context, the experience, the evaluation, the human understanding of what the situation actually requires — is the value that no AI augmentation eliminates.
Second: identify one area of your work where AI tools could genuinely help you — not replace you, but free you from the more mechanical parts so you can spend more time on the aspects that require your actual human judgment.
Write what that looks like. What would it free you to do more of?
That is the practical partnership. You take the parts that require a human. The tools take the parts that don't. And what you produce together is more than either could produce alone.
The most useful frame for AI is not 'threat' or 'saviour.' It is 'tool.'
A powerful, novel, sometimes unreliable, rapidly developing tool — that, used well and with human oversight, expands what you can do.
Your job is not to compete with it. It is to direct it. To evaluate it. To add what it cannot provide. To remain the human in the loop — the person whose judgment and care and accountability determine what the tool's outputs actually mean and what to do with them.
That job is not going away. It is becoming more important.
I'll see you in the final lesson.