The Two-Year Shakeout: Which Jobs Are Disappearing and Which Are About to Be Worth More Than Ever

Predicting the future of jobs is a reliable way to be wrong. Experts have been forecasting mass unemployment from automation since the 1960s, and the unemployment rate has stubbornly refused to cooperate. So let me be clear about what this article is and isn't.

This isn't a prediction that millions of people will be unemployed in two years. That's almost certainly not going to happen. What is going to happen — what is already happening — is a rapid reshuffling of which skills the market pays for. Some roles will compress, merge, or evaporate. Others will expand, command higher rates, and become harder to fill.

The distinction matters. The total number of jobs isn't shrinking. The specific jobs that exist are shifting. And two years is enough time for that shift to be very real and very painful for people on the wrong side of it.

The Roles That Are Compressing

Let me be precise with language. "Disappearing" is too strong for most of these. What's actually happening is compression — the same output being produced by fewer people, or the role being absorbed into an adjacent one. The job title might persist, but the number of humans doing it shrinks significantly.

Entry-Level Copywriting and Content Production

This is already well underway. The junior copywriter who was hired to produce blog posts, product descriptions, email drafts, and social media captions is being replaced by a senior marketer with AI tools. Not because the writing doesn't matter — it does — but because the production of competent first drafts is no longer a specialized skill. A marketing manager who previously oversaw three junior writers can now do the work herself with AI, faster, and often better because she understands the strategy behind the writing.

The writers who survive — and many will — are the ones who bring something AI can't: a distinctive voice, subject matter expertise, investigative skills, or the ability to write in a way that makes people feel something. The commodity layer of writing — the "we need 20 blog posts about kitchen faucets" layer — is compressing fast.

Data Entry and Basic Data Processing

Any role whose primary function is moving data from one format to another, cleaning spreadsheets, categorizing information, or generating standardized reports is on an extremely short timeline. AI handles these tasks with near-perfect accuracy, at a fraction of the cost, and without getting tired at 3 PM on a Friday. This isn't a prediction. It's a description of what's already happening in accounting firms, insurance companies, and administrative departments across every industry.

First-Tier Customer Support

The chatbot of 2022 was a frustrating menu tree that made you type "speak to a human" in all caps. The AI support agent of 2026 can understand nuanced questions, access account information, troubleshoot technical issues, process returns, and handle the vast majority of customer inquiries without human intervention. The remaining human support agents handle escalations — the complex, emotional, or unusual cases that AI can't resolve. This means fewer agents total, but the ones who remain are more skilled and higher-paid.

Junior Graphic Design

The entry-level designer who was hired to resize social media graphics, create simple layouts from templates, and produce variations of existing designs is being displaced by AI tools that do the same work in seconds. Canva's AI features, Midjourney, and a dozen other tools have made basic visual production a commodity. The design profession isn't disappearing — but the entry ramp into it is changing. The junior role that used to be "production assistant doing repetitive layout work" is evaporating. What remains requires actual design thinking, brand strategy, and creative direction.

Basic Translation

AI translation has reached the point where it handles most standard business and informational translation with high accuracy. The human translator working on user manuals, website localization, and business correspondence is being displaced rapidly. Literary translation, legal translation, and any context requiring deep cultural nuance still needs humans — but the volume of basic translation work available to human translators is shrinking fast.

Bookkeeping (Routine)

The bookkeeper whose job is categorizing transactions, reconciling accounts, and producing standard financial statements is being automated out of most small businesses. AI-powered accounting tools like QuickBooks and FreshBooks are approaching the point where they handle 90% of small business bookkeeping without human intervention. The remaining 10% — the judgment calls, the unusual transactions, the strategic advice — is the accountant's job, not the bookkeeper's.

Paralegal Research

Junior paralegals who spend their days reviewing documents, searching case law, and summarizing depositions are watching AI tools eat into their workload rapidly. Legal AI can process thousands of documents in hours, identify relevant precedents, and flag potential issues with an accuracy that matches or exceeds a tired human working through their second box of files at midnight. The paralegal role won't disappear, but it will shrink in headcount and shift toward tasks that require legal judgment rather than legal research.

The Roles That Are Becoming More Valuable

Here's where it gets interesting. For every role that's compressing, others are expanding — often in unexpected ways. These aren't speculative. The demand increases are visible right now in job postings, freelance rates, and salary surveys.

AI Workflow Designers

Someone needs to figure out how AI tools fit into existing business processes. Not the tools themselves — the implementation. Which tasks should be automated? Which should be AI-assisted? Which should remain fully human? How do the handoffs work? What are the failure modes? This role didn't exist two years ago and it's already in high demand. It's part consultant, part systems thinker, part change manager. The people who can walk into a business, understand its operations, and design an AI-augmented workflow that actually works are charging premium rates and getting more work than they can handle.

Senior and Strategic Creatives

When AI handles production, the people who direct production become more important, not less. The creative director who decides what the brand should say, how it should look, and what story it should tell. The senior copywriter who can write a brand manifesto that makes people cry. The art director who makes the visual choices that give a brand its identity. The strategic layer of creative work — the taste, the judgment, the vision — has become more valuable precisely because the execution layer has become cheap. Companies no longer need ten designers to produce their marketing materials. They need one great designer to direct the AI that produces them. That one great designer is worth more than the ten ever were.

Skilled Trades

Here's the one that surprises people. Electricians, plumbers, HVAC technicians, welders, carpenters — skilled tradespeople are becoming more valuable in the AI era, not less. The reason is simple: AI can't snake a drain. It can't rewire a panel. It can't install ductwork. The trades are physical, contextual, and judgment-intensive in ways that resist automation completely. At the same time, AI is funneling more young people toward desk jobs and digital work, shrinking the pipeline of new tradespeople. Fewer people entering the trades plus steady or growing demand equals higher wages. A master electrician in 2028 may outearn many white-collar professionals.

Healthcare Providers (Direct Patient Care)

Doctors, nurses, physical therapists, mental health counselors — anyone who provides hands-on patient care is becoming more valuable. AI is transforming healthcare administration, diagnostics, and research, but the act of caring for a patient is stubbornly, beautifully human. And as the population ages, demand for direct care is growing faster than supply. AI will make healthcare providers more effective by handling paperwork, assisting with diagnoses, and personalizing treatment plans. But it won't replace the nurse who holds a patient's hand, the therapist who reads between the lines, or the doctor who makes the judgment call in an ambiguous situation.

Teachers and Educators (Adaptive)

This one has a caveat: teachers who lecture and test are being displaced by AI tutoring systems that can explain any concept, at any level, with infinite patience. But teachers who mentor, motivate, socialize, and guide — the ones who understand that education is as much emotional as intellectual — are becoming more important. The teacher who notices a kid is struggling at home. The professor who inspires a student to change their major. The trainer who reads the room and adjusts on the fly. AI makes content delivery free. It makes the human elements of teaching priceless.

Compliance and Risk Specialists

As AI gets deployed across industries, the regulatory and risk landscape is becoming more complex, not less. Who's liable when AI makes a bad decision? How do you audit an AI system for bias? How do you ensure compliance with regulations that are being written in real time? Compliance officers, risk managers, and regulatory specialists are in growing demand because the technology is outpacing the rules, and every company using AI needs someone who understands the legal and ethical terrain.

Personal and Relationship-Driven Services

Financial advisors, executive coaches, therapists, real estate agents (the good ones), wedding planners, interior designers — any role where the primary value is understanding a specific human being's needs and guiding them through a complex decision. AI can provide information. It can analyze options. It can even generate recommendations. What it cannot do is build a relationship with a person, understand their unspoken concerns, navigate their emotions, and earn their trust over time. The services built on human relationships are becoming more valuable as everything else becomes automated.

The Pattern

If you step back and squint, a clear pattern emerges. The roles that are compressing are the ones defined by repeatable execution — producing standardized outputs from standardized inputs. Write a blog post. Categorize these transactions. Translate this document. Resize this graphic.

The roles that are becoming more valuable are the ones defined by judgment, relationships, and physical presence. Decide which blog post to write and why. Advise this client on their financial strategy. Fix this electrical panel. Help this student find their direction.

AI is exceptionally good at execution. It is not good at judgment, relationships, or showing up physically. The market is repricing accordingly.

What This Means for You

If your job is primarily execution — if you could describe your daily tasks as a series of repeatable steps — you have about 18 to 24 months to evolve. Not to panic. To evolve. That means moving up the value chain in your current role (from producing content to directing content strategy), moving laterally into a judgment-heavy role (from bookkeeper to financial advisor), or building something of your own where the full spectrum of your knowledge creates value that execution alone never captured.

If your job is primarily judgment, relationships, or physical skill — you're in a strong position that's about to get stronger. The key is to adopt AI tools that amplify your capabilities rather than resisting them. The financial advisor who uses AI to analyze portfolios and generate reports can serve more clients at a higher level. The electrician who uses AI to estimate jobs and manage scheduling can run a more profitable operation. The tools make the human more valuable, not less.

And if you're in between — a mix of execution and judgment — the next two years will determine which side of the line your role falls on. Push toward judgment. Push toward relationships. Push toward the things that AI makes more valuable, not the things it makes cheaper.

The Bottom Line

Two years from now, the economy won't have fewer jobs. It will have different jobs. The roles defined by repeatable execution are compressing — rapidly and irreversibly. The roles defined by human judgment, human relationships, and human presence are expanding.

The shakeout isn't coming. It's here. The question isn't whether your work will be affected. It's whether you'll be on the side that's rising or the side that's compressing.

You still get to choose. But the window for choosing comfortably is about 24 months. After that, the market chooses for you.