I work in AI. I help companies implement it. I see the potential, the productivity gains, the efficiency, the genuinely useful applications. But I also see a lot of chaos around the human cost we're not talking about enough: this is happening right now, not in some distant future.

A 24-year-old developer told a reporter he's "basically a proxy to Claude Code." Customer service roles disappear by the thousands. Administrative teams get "rebalanced." This is invisible displacement, knowledge workers with degrees watching $50-80K careers evaporate while being told to "upskill."

Employment for developers aged 22-25 dropped 20% since late 2022. One 2023 CS grad applied to 5,762 jobs. Zero offers.

What We Say Is Available

Salesforce committed to training 16 million by 2030. Cisco and SAP pledged 25 million and 12 million. The U.S. proposed an AI Workforce Research Hub.

Plans for 2030. Displacement is happening in 2026.

What Actually Works (And Its Limits)

Workforce retraining programs show displaced workers do see increased earnings. But those retraining for high-AI-exposed jobs earn 25-29% less than those entering low-exposure roles.

Microsoft Research identified high-exposure roles: interpreters, writers, PR specialists, sales reps, customer service, administrative assistants. Brookings found 6.1 million workers face both high exposure AND low ability to transition, "skills are less transferable and reemployment prospects narrower."

Research shows low-exposure roles exist (healthcare support, skilled trades), but telling someone with a degree and ten years' experience to become an electrician isn't realistic. What matters: higher-income workers have better "adaptive capacity"—savings, transferable skills, networks. Use those advantages.

What I'm Seeing On The Ground

As Professional Development Chair for NSBE Boston Professionals, I run AI literacy sessions. Rooms are packed. Someone asked: "I understand ChatGPT. But how does this help me keep my job?"

Honest answer: I don't fully know yet. Nobody does.

We're expecting people to bridge a massive gap on their own time and money while wondering if their role will exist next year.

What You Can Actually Do This Week

Document what AI can't see. Write down why you make decisions. "I escalated this because..." That context is valuable.
Spend 30 minutes with AI on real work. See where it fails—that's where you're needed.
Talk to your manager. Ask: "How is AI changing our work? Where do you see my role evolving?"
Look for adjacent roles. Customer service → customer success, account management, user research. Administrative work → operations roles. Don't abandon your experience—shift where it's applied.
Find your community. Professional organizations, industry meetups where people discuss AI's impact in your field. Free platforms exist, but you need community to make sense of them.
Use your domain expertise. AI can't understand why your company's procurement has seventeen approval steps. Customer service pros who know edge cases become AI quality reviewers. Junior devs who understand system design review AI-generated code.

What Companies Should Do

Companies handling this well create bridge roles i.e. 6-12 month positions combining existing knowledge with new skills. Customer service lead becomes AI quality specialist. Junior dev transitions to code review.

They run "AI + Your Role" workshops where teams figure out what AI means for actual daily tasks. They build mobility programs with concrete examples: "Sarah went from customer service to AI trainer in 8 months."

Most companies call it "rebalancing" and hope the problem solves itself.

The Thing Nobody Wants to Say

We don't have this figured out. We have pilot programs and government proposals. We have people whose jobs are changing now and help arriving in 2030.

The transition is possible but requires companies to actually invest in people, not issue press releases. It requires honest conversations, not cheerful LinkedIn posts about "exciting opportunities."

After NSBE workshops, I see relief when people understand prompts. Then fear when they realize that's 1% of staying relevant.

The question isn't whether AI changes jobs. It's whether we help people through that change or pretend it's not happening.

If your job is changing, you're not imagining it. If one workshop isn't enough, you're right.

Start this week. Document your context. Use AI on real work. Talk to your manager. Find adjacent roles. Find community.

That's not a solution. But it's a start.