The Actual Numbers — Not the Headlines

Goldman Sachs Research has put hard figures on what AI will do to the US labor market, and the doomers are going to be disappointed.

Joseph Briggs, who co-leads the Global Economics team at Goldman Sachs Research, estimates 6-7% of the US workforce will be displaced during the AI transition period. That's the baseline. Under different assumptions, it could range from 3% to 14%.

The unemployment rate impact? A half-percentage-point rise above trend. Temporary — historically, technology-driven displacement disappears after roughly two years, according to the Goldman Sachs Research report.

Amazon, Meta, and Microsoft have each blamed thousands of layoffs on AI over the past year. But Goldman's own economists — according to BBC News — believe the tech sector's employment share dropping below long-term trend is the most visible impact so far. Everywhere else? No significant AI-led shift has shown up in US labor data yet.

The Upside Nobody's Leading With

AI is accelerating business creation at a measurable pace, especially among younger Americans.

Gusto — an HR and payroll platform used by half a million small businesses — surveyed 1,000 people who opened businesses in 2025. A majority said AI made the process significantly faster or cheaper, according to NPR.

Aaron Terrazas, an economist at Gusto, told NPR that barriers like hiring lawyers and administrative staff — previously luxuries for startups — are disappearing. AI handles the intake. The founder handles the vision.

Justus Shaw, a 20-something entrepreneur in Nashville, built Shaws Coffee Cart using Claude AI as a $20-a-month advisor for taxes, legal questions, and employee management. Without it, he told NPR, he would have been buried in research or expensive consultants. With it, he grew faster.

Gen Z isn't just losing jobs to AI — they're using AI to become the boss.

What Mainstream Coverage Is Getting Wrong

Most major outlets are running two parallel, contradictory narratives without connecting them.

One story: AI is destroying jobs. Cue the Amazon and Meta layoff announcements. Cue the anxiety pieces about computer programmers and graphic designers.

The other: AI is creating a boom. Cue the Silicon Valley investment stories.

Both are partially true. Neither is the full picture.

Goldman Sachs Research identifies the real displacement risk clearly: computer programmers, accountants and auditors, legal and administrative assistants, and customer service representatives face the highest exposure. These are real people with real careers, and pretending otherwise is dishonest.

But Goldman also projects that AI will raise US labor productivity by around 15% when fully adopted. That's enormous. And the infrastructure buildout alone — data centers, power grids — will require roughly 500,000 net new jobs in the US by 2030, according to Goldman analyst Evan Tylenda.

Electricians. HVAC contractors. Lineworkers. Construction crews. Already, Goldman reports that hiring for these roles is running above trend relative to the long-term average.

The Raspberry Pi Warning Nobody Should Ignore

Eben Upton, founder of Raspberry Pi — the British computer maker that has put low-cost computing in the hands of millions of students globally — issued a direct warning to BBC News this week.

Overestimating AI's ability to replace human workers could discourage young people from pursuing tech careers, Upton said. That makes the existing tech skills shortage worse, not better.

"You read in the paper: 'What guidance should you give your child about what GCSEs to choose in the context of an AI future?' We have no data to inform a rational decision," Upton told BBC's Big Boss Interview podcast.

The hype warps real decisions by real families. A teenager who decides NOT to learn to code because "AI will do it" is making a career choice based on speculation, not data. And Goldman Sachs Research's own timeline puts wide-scale AI adoption at roughly 10 years out — not next semester.

The Numbers

• 6-7% workforce displacement over a decade — Goldman Sachs Research
• Unemployment rate up half a point above trend, temporarily — Goldman Sachs Research
• 25% of all US work hours potentially automatable by task — Goldman Sachs Research
• 300 million jobs globally exposed to AI automation — Goldman Sachs Research
• 500,000 new skilled jobs needed in US energy/infrastructure by 2030 — Goldman Sachs Research
• Majority of 2025 startup founders say AI made their launch faster and cheaper — Gusto survey via NPR
• Productivity gains of 15% projected when AI is fully integrated — Goldman Sachs Research

This is NOT the apocalypse. It is also NOT nothing. It is a major economic transition that will hurt specific groups — particularly white-collar knowledge workers in mid-skill roles — while creating real opportunity for people willing to adapt.

What This Means for Regular People

If you're in customer service, legal admin, basic accounting, or entry-level programming, the threat is real and the timeline is shorter than you'd like. Waiting isn't a strategy.

If you're young and considering starting a business, the barriers just dropped dramatically. The tools exist. The cost is low. Gen Z is already moving.

And if a newspaper is telling your kid to reconsider studying computer science because AI will take all those jobs — Eben Upton wants you to know that's speculation dressed up as guidance. Treat it accordingly.