The Middle Class in the Early AI Era

The other day, I stumbled upon a comforting yet jarring realization: I am not the only one feeling completely lost right now.

We keep hearing that AI is going to change the world. For corporate professionals like us, that excitement comes with an uncomfortable question: Where will we fit into it?

For those of us whose lives depend on a salary, this is more than an interesting debate about technology. There are bills to pay, commitments to keep, and careers we have spent years building.

So how do we move forward without putting everything at risk?

Three possible paths

When I look at the choices in front of us, three paths stand out.

1. Go all-in: Leave the corporate job and commit fully to learning AI or building an AI-driven business.

2. Moonlight: Keep the day job while setting aside time outside work to learn and experiment.

3. Double down: Invest more deeply in the current career, take ownership of important work, and strengthen the value we bring to the company.

Each path involves effort and trade-offs. Doubling down is an active choice, and it can make sense. But becoming valuable in today's organization does not guarantee that our role will stay the same. We still need to pay attention to how the work is changing.

The idea that helped me think about this came from Graham Weaver's talk, “How to Live an Asymmetric Life.”

Thinking about the upside and downside

What stayed with me was the idea of looking for opportunities where the potential upside is much larger than what we could lose. The downside can be reduced, but it cannot simply be wished away.

Here is my own simplified example. Imagine investing $1 million in an opportunity that could eventually be worth $3 million, while planning to limit your loss to $200,000.

That sounds attractive. But how likely is the upside? How long might it take? And can you actually keep the loss within that limit? A planned cap is not a guarantee.

The numbers give us a starting point for judgment. We still have to understand the risk.

I think the same approach is useful when considering AI. Going all-in may open up possibilities, but leaving a dependable income behind can create serious pressure. Staying employed gives us room to learn, provided we actually make use of that room.

The question then becomes: How much time, energy, and money can we realistically invest in building new capabilities?

The uncomfortable part of learning

For an experienced professional, starting again can feel awkward.

We may be confident in our field and suddenly find ourselves struggling with something unfamiliar. It can be difficult to separate a reasonable concern about a new tool from the discomfort of no longer feeling like the expert.

Some skepticism is useful. We should question what AI can do and where it falls short. But if we reject it before exploring it, we may never discover where it could help us.

We have to give ourselves permission to be beginners in one area while remaining experienced in another.

Then comes the less exciting part: repeated effort.

I think of building muscle. Progress takes regular training and enough recovery. One exhausting session does not create lasting strength. Learning a new way of working takes that same willingness to return, practise, and improve over time.

There are real trade-offs. A few hours spent learning have to come from somewhere. Paying for a useful AI subscription may mean spending less on another discretionary purchase.

The challenge is to make those choices deliberately and find a pace we can sustain alongside work, rest, and the people who matter to us.

When does going all-in make sense?

But what if we are already learning, experimenting, and building something of our own?

At some point, the question may become: Should I leave my corporate job and commit fully to this business?

This is where I want a clear measure for myself.

First, I would calculate my total annual corporate compensation: salary, benefits, and bonuses. That is the value I would be walking away from.

My personal threshold is conservative: before I would consider going full-time, the AI business would need to consistently generate personal income available to me after business expenses of at least five times my current corporate compensation, compared on the same annual basis.

That is my 5x threshold. It is a personal decision rule, and hitting it would still leave risks to assess. Other people will have different responsibilities, resources, and levels of risk they can accept.

For me, keeping the day job while building toward that point protects an important source of stability and gives me time to develop the business.

The question for tomorrow morning

This brings me back to the office.

If time is already our most constrained resource, learning only in the evenings and on weekends may be difficult to sustain. There is another possibility worth exploring: building AI capability through the work we already do.

So the question I want to take into my next working day is:

“How can I integrate AI into my current eight-hour workday in ways that help me contribute today, become more adaptable, and create more choices for my future?”

In part two, I will explore what that integration could look like in practice.