How AI Is Fueling Growth While Leaving Millions Behind

Growth Is Real. So Is the Gap.

The US economy grew at a 2.1% annualized rate in the first quarter of the year, driven in meaningful part by businesses ramping up AI-related investments. That’s a real number. It reflects real activity. But aggregate growth has a way of flattering the picture.

Consumer sentiment is hovering near record lows. The bottom quarter of Americans on the income spectrum have seen the weakest wage growth of any cohort this year. And demand at the Richmond Neighborhood Center’s food pantry is up roughly 10%.

These aren’t contradictory data points. They’re the same story told from different zip codes.

The Geometry of the Boom

The AI investment surge has created a specific geography of winners. Tech hubs—San Francisco, New York, Seattle, Los Angeles, San Jose, Washington DC—have absorbed the bulk of the capital and the talent. San Francisco companies alone account for nearly two-thirds of worldwide AI funding, according to data firm Crunchbase.

That concentration of capital creates a cascade effect. High salaries flow to a relatively small group of workers. Those workers spend. Housing prices respond. Rents climb. And suddenly, the economic growth that looks healthy from 30,000 feet is actively making life harder for the people already struggling at street level.

The wealthiest 10% of Americans are now responsible for as much as 62% of US economic growth, according to Moody’s. That’s not a sign of a broadly thriving economy. That’s a sign of an economy running on a very narrow engine.

Winners, Losers, and the People Writing the Training Data

Manuel Pastor, director of the Equity Research Institute at the University of Southern California, put it plainly: “You’re seeing incredible concentrations of wealth as a result of AI for these new companies, their founders and their first employees. It’s exacerbating an economy of winners and losers.”

The winners are identifiable. Early investors. Founders. First employees at companies that are now worth billions. SpaceX debuted on Wall Street as the largest IPO on record, now valued at more than $2.1 trillion. OpenAI and Anthropic, both headquartered in San Francisco, are reportedly preparing for their own IPOs—moves that would add trillions in new market value and reward a specific slice of the population handsomely.

The losers are harder to photograph but easier to find. Recent college graduates struggling to break into a job market that’s increasingly automated at the entry level. Low-income Americans carrying debt while inflation continues to bite. And—this one tends to get underreported—creative workers.

“What people put on the internet or put into books is being privatized by these AI companies, making it more difficult for those same people to make money,” Pastor said. “That’s happening to people who are authors, to people who are musicians, anyone who is a creative.”

There’s a particular irony in that dynamic. The content that trained these models came from somewhere. The people who made it are now competing with the systems built on their work.

Main Street Is Getting Squeezed Too

The inequality story isn’t just about individuals. It’s reshaping business investment patterns in ways that should concern anyone paying attention to the broader economy.

Maxime Darmet, senior economist at Allianz Trade, noted that if you strip out AI-related spending, business investment would actually be falling—something that typically only happens during recessions. “The technology is powerful in propping up the economy, but at the same time, there’s a lot of spending being cut in more traditional areas.”

That’s a structural shift, not a blip. Capital that might have flowed into manufacturing, retail infrastructure, or local services is being redirected toward AI development and deployment. The aggregate GDP number stays healthy. The Main Street business owner wondering why foot traffic is down doesn’t feel that health.

San Francisco as a Mirror

Yves Xavier, community programs director at the Richmond Neighborhood Center, is careful not to draw a straight causal line. “We can’t draw a direct line to AI’s impact and say ‘That’s exactly it’ because it’s been happening for a while,” he told CNN. “But it doesn’t exactly take a rocket scientist to see how that’s widening the inequalities in a city already dealing with those issues.”

That’s an important nuance. San Francisco’s inequality predates the current AI boom. The tech industry has been reshaping the city’s economics for decades. What AI has done is accelerate and amplify dynamics that were already in motion—more capital, higher salaries, faster displacement, steeper rents.

The food pantry waitlist of 200 people isn’t a new phenomenon. It’s an old phenomenon getting worse in a city that’s simultaneously becoming the financial capital of the most heavily funded technology sector in history.

What This Means for the Tools Ecosystem

For anyone operating in the AI tools space—building products, evaluating platforms, making software decisions—this context matters more than it might seem.

The AI tools market is not a neutral playing field. It reflects the same concentration dynamics visible in the broader economy. The best-funded platforms attract the best talent, the most integrations, the fastest iteration cycles. Smaller tools built for industries or communities outside the tech hub orbit often struggle to compete for attention, capital, or distribution.

That doesn’t mean the tools themselves are bad or that using them is ethically fraught. But it does mean that the “AI is democratizing everything” narrative deserves scrutiny. Access to AI tools is not evenly distributed. The ability to benefit from them—through training, infrastructure, reliable internet, and the time to experiment—is not evenly distributed either.

The businesses and workers most likely to be disrupted by AI are often the least equipped to adopt it quickly enough to stay ahead of that disruption. That’s not a technology problem. It’s an economic structure problem that technology is currently making more visible.

The Uncomfortable Arithmetic

Here’s the tension that doesn’t resolve cleanly: the AI boom is genuinely producing economic growth. The investments are real. The productivity gains are real. The wealth being created is real.

And the food pantry waitlist is real. The stagnant wages at the bottom of the income distribution are real. The creative workers losing income to systems trained on their own output are real.

These things coexist. The economy can grow and leave people behind simultaneously—it has done so before, and it’s doing so now with unusual clarity.

The question worth sitting with isn’t whether AI is good or bad. It’s who captures the value it creates, who absorbs the disruption it causes, and whether the gap between those two groups is something policy, business practice, or market forces will eventually close—or continue to widen.

The Takeaway

If you’re evaluating AI tools for your business or workflow, the macro picture here is context, not a reason to pause. Adoption is happening regardless, and falling behind on useful tools has real costs.

But the smarter framing is to ask not just “does this tool work?” but “who does this tool work for?” The platforms and products that will hold up over time are the ones solving real problems for real users—not just optimizing for the metrics that look good in a funding deck.

The AI boom is producing winners. Being intentional about which tools you adopt, and why, is one of the few levers most people actually have in this economy.