📰 AI News: The "Permanent Underclass" Warning (And Why The Bubble Makers Might Be Just As Replaceable)

🔥

Oct '25 • AI News

A sobering prediction is making waves: AI will create a "permanent underclass" of workers left behind by automation. But here's the twist—the people building the AI bubble might be just as replaceable as the rest of us.

The doomsday prediction:

According to The New Yorker, OpenAI employee Leopold Aschenbrenner claims AI will "reach or exceed human capacity" by 2027. Once it develops the capacity to innovate, AI superintelligence will supersede even the need for its own programmers—and then wipe out jobs for everyone else.

Nate Soares, co-author of the AI critique If Anyone Builds It, Everyone Dies, suggests "people should not be banking on work in the long term."

Math tutors, cinematographers, brand strategists, and journalists are all cited as panicking about their futures.

The uncomfortable reality check:

Guardian columnist Van Badham offers a brutally honest take: If you're panicking about being forced into the permanent underclass, you're probably already in it.

Inherited wealth creates more billionaires than entrepreneurship. The opportunity gap is growing. If your family can't fund your tech startup or media empire, it's likely because they were already in a tech-displaced underclass.

The historical pattern:

This isn't new. Technology has been displacing workers for generations:

→ Spinning jennies replaced textile workers→ Combine harvesters replaced farm laborers→ Electrification changed entire industries→ Trains, cars, and recording technology devastated previous job markets→ Computers automated office work→ Apps replaced betting shops→ iPhones replaced admin assistants

Every generation faces technological displacement. The question isn't if it happens—it's how we respond.

The union solution (it's worked before):

Japan, 1970s: When automation hit carmaking, unions successfully fought for "cooperative modernization" with job security guarantees and lifelong retraining programs.

Australia, 1980s: As technology replaced nursing functions, nurses fought to enhance their status through tertiary training with new technologies—improving both work conditions and professional standing.

Writers Guild of America, 2023: Strike resulted in an agreement that writers wouldn't be replaced by generative AI as source material or be compelled to use it—but could use AI as a productivity tool at their discretion.

The pattern: Organized labor has successfully negotiated technological transitions before. It can happen again.

The bubble reality (AI isn't inevitable):

Despite the hype from tech zealots, AI faces serious sustainability problems:

💸 Financial math doesn't work - Harris Kupperman from Praetorian Capital calculates AI data centers may depreciate in just three years. "No one understands how the financial math is supposed to work."

📊 Investors are worried - Majority of Bank of America survey respondents admit concern that AI stocks are in a bubble

🏦 Central banks are concerned - Bank of England warned of potential "sudden correction" in worldwide markets due to soaring AI company valuations

🔬 MIT research shows failure - 95% of corporate clients trialing AI pilots or initiatives find zero or negligible return

❌ Carnegie Mellon findings - AI agents failed to complete basic office tasks about 70% of the time

😬 Companies regret AI layoffs - 55% of surveyed companies that replaced workers with automation now regret the decision

💰 Layoffs cost more than they save - Analytics firm Orgvue found layoffs cost companies $1.27 for every $1 saved

The theft problem:

Nick Clegg (professional tech spokesperson) claimed that paying copyright on AI's stolen content would "basically kill" the industry.

Translation: If AI companies had to actually pay for the content they exploit, the business model collapses.

As Badham puts it: "Not paying for a product you exploit to sell your product is theft; not paying for labor you employ is slavery. If your business model is theft and slavery, you're not an industry. You're a pirate."

The delicious irony:

Badham used AI to create a "techbro bot" that generates social media posts as a tech lord, informed by public statements from actual tech leaders.

The result? The AI-generated tech hype is "indistinguishable, digitizable—and replaceable."

In other words: The makers of this bubble are just as replaceable as everyone else.

Why this matters:

⚠️ The threat is real, but not inevitable - AI displacement is happening, but historical precedent shows organized response can shape outcomes

🤝 Collective action works - Unions have successfully negotiated technological transitions before, securing retraining, job protections, and enhanced roles

💭 The hype exceeds the reality - Despite apocalyptic predictions, current AI still fails at basic tasks 70% of the time and delivers zero ROI for 95% of companies testing it

💰 The economics are shaky - When your business model requires not paying for content or labor, and data centers depreciate in three years, sustainability is questionable

🎭 Even the "irreplaceable" are replaceable - AI can already replicate tech leader rhetoric and hype convincingly, suggesting the people building AI aren't as special as they think

What this means for businesses:

🛡️ Don't panic, prepare strategically - Yes, AI will change jobs. But most companies implementing AI see zero return. You have time to adapt thoughtfully, not reactively.

👥 Invest in your people, not just technology - 55% of companies that replaced workers with AI regret it. Your competitive advantage is skilled humans using AI as a tool, not AI replacing humans.

⚖️ Consider the ethical and legal risks - If AI's business model is built on unpaid content and labor, regulatory correction is coming. Companies that build sustainable, ethical AI practices will survive what's coming.

📚 Learn from history - Every technological revolution created displacement AND new opportunities. The winners were those who adapted roles, not those who clung to obsolete ones or blindly replaced humans.

🔧 AI as tool, not replacement - The Writers Guild model is instructive: AI as a productivity tool at worker discretion, not as a replacement for human expertise and judgment.

💡 Question the ROI before investing - If 95% of companies see zero return on AI pilots, don't assume you'll be in the 5%. Test carefully, measure religiously, and be honest about results.

🤖 The "permanent underclass" isn't predetermined - How we respond to AI displacement—through policy, organizing, retraining, and strategic adoption—will determine outcomes more than the technology itself.

The bottom line:

Yes, AI threatens jobs. But so did electricity, cars, and computers—and we're still here.

The difference between becoming part of a "permanent underclass" and thriving in an AI-augmented economy isn't the technology. It's how we respond: collectively, strategically, and with clear eyes about both AI's capabilities and its very real limitations.

And here's the kicker: If AI can already replicate the hype and rhetoric of the tech leaders building it, maybe the "permanent underclass" isn't where you think it is.

Maybe we're all in this together. Which means we need to navigate it together—not as passive victims of inevitable change, but as active participants shaping what comes next.

Your take: Does this prediction make you more motivated to learn AI as a defensive skill, or does it reinforce that human skills AI can't replicate are where the real value is? Drop your honest reaction below. 🤔

7 comments