The Underclass - Part IV: The Good, The Bad, The AI
The New Underclass: Technology, Crisis, and Class Mobility
Technology is annoying.
Not because it’s bad, but because it’s never just one thing.
Like the Industrial Revolution before it, digital technology has been both a ladder and a trap. It has created extraordinary opportunity while quietly dismantling the structures that once made that opportunity broadly accessible.
So let’s do this properly.
The Good.
The Bad.
And then… the Ugly.
The Good: When the Internet Actually Worked
Here’s the part that makes this whole thing frustrating.
Technology genuinely did expand opportunity.
The internet, in particular, opened more career paths for more people than any previous technological shift in human history. Entire industries appeared out of thin air. Barriers collapsed.
My entire business would not exist without it.
Websites. E-commerce. Online services. Digital products. Remote work. Creative industries that didn’t require gatekeepers, offices, or permission slips.
Knowledge was no longer locked behind universities or private libraries. Anyone with a connection could learn almost anything. Wikipedia alone flattened centuries of informational hierarchy. The fact that nearly all the research for this article came from the internet says everything.
The barriers weren’t gone, but they were dramatically lower.
There’s actual evidence to support this optimism. Longitudinal studies have shown that internet use correlates with increased social class mobility, even after accounting for age, education, and background. Information access mattered. Skill acquisition mattered.
Globally, the effect was even more pronounced.
In developing countries, mobile phones and internet access empowered small entrepreneurs and farmers. Market information improved income. Mobile banking created savings and investment pathways. Remote work connected people in poorer regions to wealthier markets.
The tech industry itself became a mobility engine. During the dot-com boom and beyond, people from modest backgrounds became software developers, founders, and engineers, jumping directly into the upper middle class or beyond.
Social media added entirely new careers:
content creators,
YouTubers,
app developers,
esports players,
UGC creators.
E-commerce allowed small businesses to punch far above their weight. Formal education wasn’t always required. Gatekeeping weakened.
For a while, it really did feel like anyone could build something.
The Bad: When the Utopia Started Crumbling
Then the distribution problem showed up.
Those benefits were never evenly shared.
The first crack was the digital divide.
Without reliable internet, devices, and digital literacy, all that opportunity evaporates. Rural and remote communities. Lower-income households. Older workers. They didn’t get the same returns.
Even in Australia, a country with relatively strong infrastructure, access is uneven. In other parts of the world, it’s far worse. Which means talent still migrates toward connectivity, draining regions and reinforcing inequality.
A new class dimension emerged: digital capital.
Not just money.
Not just education.
But access, literacy, and algorithmic visibility.
Then came the second crack: winner-take-all economics.
The internet didn’t just democratise markets, it concentrated them.
Anyone can start an online business. But you’re competing against:
millions of others doing the same thing,
plus multinational platforms with near-infinite capital,
plus algorithms designed to amplify incumbents.
The biggest players get bigger. Everyone else gets diluted.
Big Tech created unimaginable wealth, but it flowed primarily to founders, early investors, and long-term insiders. That’s not morally shocking; that’s how capital works. But the scale of it matters.
The trajectory of growth itself became the problem.
Meanwhile, traditional middle-class jobs were quietly dismantled. Journalism. Retail. Media. Travel agencies. Entire professions were either eliminated or converted into precarious, gig-like work.
The Ugly: When AI Enters the Room
And then, of course, comes AI.
The internet disrupted information. AI disrupts labour itself.
Artificial intelligence can now perform tasks that were once firmly white-collar:
drafting reports,
analysing data,
diagnosing disease,
generating code,
driving vehicles.
This raises the most uncomfortable question in this entire series:
Does AI erase class mobility altogether?
Not in a sci-fi “robots take over” way, but in a quieter, more bureaucratic one.
If AI hollows out the middle, automating professional and administrative roles, what’s left?
A small group of highly paid technical and managerial elites.
A large group of low-paid service and care workers.
And a growing number of people who simply don’t fit anywhere.
Economists like Daron Acemoglu have warned that, without policy intervention, AI is being developed along an “automation agenda”, one that replaces labour faster than it creates new roles.
That’s how underclasses form.
Not through laziness, but through redundancy.
The fear now is the emergence of a semi-permanent technological underclass:
people unable to find stable work,
or trapped in roles AI can’t (yet) do,
jobs that are low-paid, insecure, and socially undervalued.
Care work. Delivery. Platform labour. Content moderation. “Human glue” work that keeps systems running but offers no future.
If your job disappears, and you don’t have access to advanced education or retraining, or simply can’t afford the time, mobility doesn’t stall. It reverses.
We start rebuilding medieval dynamics, unintentionally. Almost like history is on repeat.
Vibe Work and the Illusion of Participation
We’re already seeing early signs of this.
“Vibe workers.”
People who copy, paste, supervise, or lightly prompt AI systems. No ownership. No creativity. No security. No long-term skill accumulation.
Work that looks modern but feels hollow.
Participation without progression.
The Two-Edged Sword (Again)
And yet, here’s the most frustrating part.
AI could reduce inequality.
If harnessed properly, it could:
personalise education,
accelerate skill acquisition,
automate drudgery,
free people for creative and analytical work.
Entrepreneurship could become even more accessible. In theory, anyone could build an app overnight. Anyone could prototype, test, launch.
But we already know how this plays out.
Oversaturation. Platform dominance. Visibility controlled by algorithms.
The same pattern repeats.
Globally, the story is even more complex. Remote work has allowed people in lower-income countries to leapfrog into the global middle class. A developer in India or Nigeria can now work for overseas clients and earn life-changing income.
At the same time, that outsourcing undercuts middle-class roles in developed nations.
Opportunity expands and collapses simultaneously.
Where This Leaves Us
Technology is a massive disruptor.
You don’t get the good without the bad.
You don’t get productivity without displacement.
You don’t get opportunity without erosion.
What’s becoming impossible to ignore is this:
The middle class is thinning.
The winners are consolidating.
The floor is dropping.
Not because of ideology.
Not because of morality.
But because of structure.
And unless that structure changes, the underclass won’t be a prediction.
It’ll be policy by default.
Come back tomorrow for the final article in this series, Part V: No Clean Endings