Why Every Technology 'Kills' Jobs — Then Creates More (AI Follows the Same Pattern)

Part 2 of 3: The Pattern | By Bilal El Alamy | March 2026 ← Part 1: What the Data Actually Shows | → Part 3: The Real Disruption

---

TL;DR

• Bank teller employment grew 11.6% during peak ATM expansion (1997–2007)
• Bookkeeper employment grew 6.4% after spreadsheet adoption
• 60% of US workers today hold jobs that didn't exist in 1940
• 85% of employment growth since the industrial revolution came from technology-created occupations
• The mechanism is always the same: automate tasks → lower costs → expand demand → create new roles

---

In Part 1, I showed you the data: zero systematic unemployment increase in AI-exposed jobs, a 61-point gap between AI capability and adoption, and every major institution projecting net job creation by 2030.

But data without context is just numbers. To understand why the AI job displacement myth keeps getting the story wrong, you need to see the pattern it keeps ignoring — a pattern that has repeated with every major technology for over two centuries.

---

The Pattern That Repeats: Technology "Kills" Jobs, Then Creates More

1970s–2000s: ATMs and Bank Tellers

ATMs were supposed to eliminate bank tellers. The logic was airtight: if a machine dispenses cash, you don't need a human to do it.

What actually happened: ATMs reduced the cost of operating a branch. When branches got cheaper to run, banks opened 43% more branches in urban areas. More branches needed more people. Bank teller employment grew 11.6% from 1997 to 2007 — the peak decade of ATM proliferation.

The role evolved. Tellers stopped being cash dispensers and became relationship advisors. The job didn't die — it upgraded.

1980s–1990s: Spreadsheets and Accountants

VisiCalc launched in 1979. Excel followed. Manual ledger calculations — the core of bookkeeping — were automated overnight.

What happened: Bookkeeping employment grew 6.4% from 1997 to 2007. When accounting got cheaper and faster, more businesses could afford professional financial management. The total demand for accounting services expanded faster than automation could shrink it.

Accountants stopped being calculators and became strategic advisors. Same job title. Completely different — and more valuable — work.

1990s–2010s: E-Commerce and Retail

Amazon was supposed to kill retail. Physical stores would become irrelevant. Why drive to a mall when you can click?

Reality: there were more sales clerks in 2009 than in 1999 — three-quarters of a million more. In-store retail shifted from transaction processing to experience and consultation. Online and offline didn't substitute — they expanded the total retail surface area.

2020s: AI and Knowledge Work

Generative AI is the latest candidate for mass displacement. Three years in: no systematic increase in unemployment in exposed occupations. The Anthropic data is clear on this.

The pattern holds — again.

Sources: IMF (Bessen, 2015), BLS Occupational Employment Survey, AEI

---

The Mechanism Nobody Talks About

The mechanism is always the same:

Technology automates tasks → Cost of delivering a service drops → Demand for the service increases → New roles emerge — often better ones

This isn't magic. It's basic economics. When the cost of something falls, people buy more of it. When the cost of accounting falls, more businesses use accountants. When the cost of medical diagnostics falls, more patients get tested. When the cost of content creation falls, more content gets created.

The ATM didn't kill bank tellers — it turned them into relationship managers. Spreadsheets didn't kill accountants — they turned them into strategists. AI will do the same to knowledge workers — not by replacing them, but by elevating the work they do.

The Goldman Sachs Number Everyone Misquotes

Goldman Sachs estimated approximately 300 million tasks affected by AI. Real number. But tasks affected ≠ jobs eliminated. Every major technology transition affects tasks:

• The ATM affected cash dispensing. Not banking.
• Spreadsheets affected manual calculations. Not accounting.
• E-commerce affected in-store browsing. Not retail.

The AI job displacement myth assumes a fixed number of tasks in the economy. But technology doesn't just automate existing tasks — it lowers the cost of value creation, which expands the total economy, which creates new tasks and new roles that didn't exist before.

85% of Job Growth Came From Technology

Goldman Sachs puts a striking number on this pattern: roughly 60% of U.S. workers today are employed in occupations that didn't exist in 1940.

That means over 85% of all employment growth in the last 80+ years has come from technology-driven job creation. Not job preservation — job creation.

Web developer. Social media manager. Data scientist. Cloud architect. UX designer. Content strategist. DevOps engineer. AI trainer. Prompt engineer. None of these existed 30 years ago. Most didn't exist 15 years ago. All were created by technology that critics said would eliminate jobs.

---

Who Gets Transformed, Who Gets Ahead

"Most exposed" doesn't mean most endangered. It means most transformed.

The consistent pattern: AI handles the routine, humans handle the judgment. The job doesn't disappear — it upgrades.

This is exactly what we see in practice. At MrChief, operators use AI agents for the mechanical work — research, data processing, scheduling, triage — while keeping strategic decisions, client relationships, and creative work firmly in human hands.

---

The Skills That Survive and Thrive

WEF projects 39% of current skills will transform by 2030. That's not disaster — it's the same rate of skill transformation that every previous technology wave has caused.

Skills declining: routine data processing, basic information synthesis, manual report generation.

Skills rising: AI collaboration, systems thinking, interpersonal complexity, ethical reasoning, creative strategy.

PwC's AI Jobs Barometer found a 56% wage premium for workers who demonstrate AI skills — up from 25% the year before. That premium is a market signal. Organizations are paying up for humans who can work alongside these tools.

77% of employers plan to reskill their workforce for AI (WEF). The companies treating this as "AI enables layoffs" will lose the talent war to companies treating it as "AI enables capability multiplication."

The Fastest-Growing AI Roles Are Not All Technical

Source: Autodesk AI Jobs Report 2025, GlobalData analysis of ~3M job listings

AI Content Creator — a non-technical role — is growing nearly as fast as AI Engineer. Design has overtaken coding as the most in-demand skill in AI job postings. AI isn't creating a world where only engineers thrive. It's creating a world where every professional can leverage AI to expand their range.

---

The T-Shaped Professional Becomes Star-Shaped

The traditional career model was built on depth. You picked a lane — law, medicine, engineering, finance — and spent decades deepening expertise. The "T-shaped" professional was the upgrade: deep in one area, broad awareness of others.

AI breaks this model — in the best way.

When you can summon expert-level capability in any domain within seconds, the shape of professional competence changes. You can be deep in your core domain while deploying meaningful capability across five, ten, twenty adjacent fields.

Star-shaped professional: Deep in their core. AI-enabled competence radiating outward into legal, financial, marketing, operations, technical — whatever the work demands.

This is what MrChief's agent architecture is designed for — not to replace expertise, but to give every operator the breadth of support previously available only at enterprise scale. The security architecture ensures that breadth doesn't come at the cost of compliance.

---

What the Pattern Tells Us About 2026–2030

If history is any guide — and 200 years of data says it is — here's what comes next:

1. AI adoption accelerates — the 61-point capability gap closes, not because jobs disappear, but because workers learn to use the tools
2. New job categories emerge — just as "social media manager" emerged from social platforms, "AI workflow designer" and "agent orchestrator" emerge from AI platforms
3. Existing jobs upgrade — every profession gets more productive, which makes every professional more valuable
4. The wage premium grows — PwC's 56% premium is early. It will widen before it normalizes.

The question is never "will AI eliminate jobs?" It's always: "what new jobs will this create, and am I positioning to do them?"

---

→ Continue to Part 3: AI Capability Democratization — The Real Disruption The real shift isn't job loss. It's the dissolution of professional walls. When anyone can do anything, the question changes from "what's my job?" to "what can I do that's useful?"

← Back to Part 1: What the Data Actually Shows

---

FAQ

Q: Did ATMs actually eliminate bank teller jobs?

No. Bank teller employment grew 11.6% from 1997 to 2007, the peak decade of ATM expansion. ATMs reduced the cost of running a branch, which led banks to open 43% more urban branches, which required more tellers. The role evolved from cash clerk to relationship advisor.

Q: What percentage of today's jobs didn't exist in 1940?

60% of U.S. workers today hold jobs that didn't exist in 1940, according to data cited by Goldman Sachs. Over 85% of employment growth in the last 80+ years came from technology-driven job creation.

---

Bilal El Alamy is the Founder of PyratzLabs and Artificial-Lab. MrChief.ai is an AI Chief of Staff platform with 100+ specialized agents for founders and knowledge workers.

Sources: IMF (Bessen, 2015), BLS Occupational Employment Survey, AEI, WEF Future of Jobs Report 2025, PwC AI Jobs Barometer, Goldman Sachs, Autodesk AI Jobs Report 2025, Anthropic (March 2026).