AI at Work: What Kind of Future Are We Really Building?

Every time a big new technology shows up, we all ask the same question:

“What will this do to jobs?”

We’ve been here before.
The steam engine pulled people into factories.
Electricity powered mass production and longer workdays.
Computers automated routine office work and created entire new professions.

AI is the next wave — only this time, the change is much faster.

• Work is being redesigned for humans + AI agents working together
  
• Productivity is less about “how many people you have” and more about how well people use AI
  
• Skills, growth, and learning are all moving on different timelines

We’re past the point of asking if AI will change the labor market.
The real question now is: what are the early signs telling us—and where could this go next?

What the early data is (and isn’t) saying

Let’s start with something important:
There’s no clear evidence of an AI-driven job apocalypse so far.

• Recent research shows that overall employment has stayed pretty stable since generative AI took off in 2022.
  
• But if you zoom into the tech sector, you start to see hints of what might come next.
  

Roughly 70% of tech companies already use paid AI tools (compared to about 40% in other industries). That makes tech a kind of early warning system for the rest of the economy.

Here’s what’s happening there:

• New-grad hiring is down in AI-exposed roles
  
  • entry-level coding
    
  • basic customer support
    
• Experienced talent is in higher demand
  
  • people who know how to use, supervise, and integrate AI
    
  • companies want graduates who can perform closer to mid-level professionals
    
• High-profile tech layoffs keep hitting the news
  
  • not necessarily proof of permanent job collapse
    
  • but a reminder that big transitions often create short-term pain before new roles appear

Across the wider economy, the range of tasks AI can automate is expanding fast:

• Routine cognitive work (data entry, simple reporting, basic analysis) is increasingly at risk
  
• People who know how to work well with AI are already starting to see wage premiums

This pattern is familiar from past tech waves.
What’s new is the speed. AI is compressing changes that once took decades into just a few business cycles.

That sets up four possible futures for the labor market—depending on what we do next.

Four possible futures for jobs in the age of AI

Imagine a simple graph:

• One axis: employment (how many people have work)
  
• The other: growth (how much value the economy is creating)
  

That gives us four possible scenarios.

1. High employment and high growth

Best-case scenario: AI makes people more capable, not replaceable.

In this future, AI acts as a power tool for humans:

• Teams produce more, faster, and at higher quality
  
• Productivity gains create new demand, new industries, and new roles
  
• The number of jobs grows, even as some tasks and roles disappear
  

We’ve seen this before with electricity, the internet, and personal computers:
when you raise the ceiling on what people can build, you usually create more work, not less.

But this outcome is not automatic. It requires leaders to:

• treat augmentation (AI helping people) as the default
  
• invest heavily in training, apprenticeships, and on-the-job experimentation
  
• help workers move from “competing with AI” to directing and amplifying it
  

In this world, learning how to work with AI becomes a growth engine, not just a defensive move.

2. High employment, low growth

Safe but underwhelming: lots of jobs, not much progress.

Here, organizations adopt AI slowly and cautiously:

• Productivity rises, but gradually
  
• Most people stay employed
  
• Wages grow a little, but not dramatically
  
• The transition feels more like evolution than disruption
  

This future reduces shock and dislocation, but also leaves a lot of potential on the table.
Think of it as “we made things a bit more efficient”… instead of reimagining how work could be done.

3. Low employment, high growth

The “barbell economy”: big gains, unevenly shared.

In this scenario, AI supercharges output, but the benefits are concentrated:

• Jobs built on repetitive, predictable tasks disappear
  
  • basic data entry
    
  • simple accounting
    
  • some customer support and back-office roles
    
• The middle thins out
  
• A relatively small group of high-skill professionals capture most of the upside
  

The economy as a whole can look “healthy” on paper—strong growth, strong profits—
while many people feel left behind.

This is where you hear phrases like:

• “winner-takes-most”
  
• “hollowed-out middle class”
  
• “high growth, weak inclusion”

4. Low employment, low growth

Worst-case scenario: disruption without the benefits.

This is the least likely, but also the most damaging outcome.

Here’s what it looks like:

• Automation outpaces adaptation
  
• Jobs are lost, but productivity doesn’t truly take off
  
• GDP stalls
  
• Social trust erodes, and political polarization deepens
  

In other words: people lose work without seeing better products, services, or living standards in return. That’s a recipe for frustration and instability.

Of course, reality won’t fit neatly into four boxes.
Labor markets are just the visible surface of a deeper shift:

• how productivity is created
  
• how policy responds
  
• how much people trust institutions and each other
  

When work changes, so do:

• where people live
  
• what communities look like
  
• how secure people feel about their future
  

Handled badly, this transition could create a “lost generation”—trained for a world that no longer exists.
Handled well, it could unlock one of the most productive and inclusive periods in modern history.

Where we are now—and what happens next

Inside companies that are actually using AI, one pattern is emerging:

• Teams that use AI to augment people are seeing productivity gains
  
• Early adopters who combine AI + human talent are often hiring faster than those using AI only to cut costs
  

At the same time:

• Some recent layoffs look like bets on future AI capacity, not just reactions to current reality
  
• Growth forecasts are mixed
  
• Many organizations are still in the “testing and piloting” phase
  

In short: the curve isn’t drawn yet.

AI could help us reach:

• high growth + high employment (the best-case scenario),
  or we could drift into stagnation and frustration if skills, institutions, and opportunities don’t keep up.

So… which future do we want?

Personally, I’m cautiously optimistic.

We can get to a world where:

• AI boosts productivity
  
• people stay employed
  
• new, meaningful roles are created
  
• more people have access to tools that used to be reserved for a small elite

But that will only happen if we:

• treat human + AI collaboration as the norm
  
• invest in skills and learning as aggressively as we invest in new models and tools
  
• get businesses, universities, and policymakers pulling in the same direction

Optimism is not inevitability.
The future of jobs isn’t something AI decides for us.

The real question isn’t:

“What will AI do to jobs?”

It’s:

“What are we willing to do to create the future we actually want?”

And that, ultimately, is a skills and choices question—one we all have a role in answering.