The Next Billion Jobs

I believe we are at the start of the largest job creation event in modern history. Not in ten years. Now.

Last month I sat across from a designer who had applied to 312 jobs in three months. She heard back from four. One was an automated rejection eleven minutes after she submitted. She asked me if her career was over. She is 28.

I have spent the last year in conversations like that one. Talented people, real experience, genuine fear. Their LinkedIn feeds tell them AI is replacing designers, engineers, product managers. Every other post is someone with fifteen years of experience who cannot get an interview.

I don't share that view. I've spent enough time with the numbers to know the panic doesn't match reality.

Every major technology in the last two hundred years created more jobs than it destroyed, even during the transition period itself. This pattern has no exceptions.

When railroads arrived in the 1840s, people predicted the end of work as they knew it. Horse breeders, canal workers, and stagecoach drivers faced extinction. What actually happened was different. Railroads created whole new industries: steel manufacturing, telegraph networks, urban retail, refrigerated food transport, and tourism. They also spurred the growth of financial services to fund all of it. Direct railroad employment grew from 85,000 to over 2 million in eighty years.

Nobody in 1840 was planning a career as a telegraph operator. The technology created demand for work that was unimaginable before it arrived.

The internet followed the same pattern. In the late 1990s, people predicted it would destroy retail and media. It would put travel agencies out of business. The dot-com crash wiped out nearly a million tech jobs. The skeptics looked right.

They were watching the beginning.

Tech employment went from 1.2 million in 1995 to over 6 million today. "Social media manager" was not a job in 2005. "Cloud architect" did not exist in 2008. "Data scientist" was barely a concept before 2010. The worst moment in tech employment was the starting line for what came next.

AI data tells a different version of the same story. I watch the numbers daily.

Open engineering jobs hit a low of 38,000 in August 2023, after the ChatGPT panic and the layoffs at Meta, Amazon, Google, Microsoft. As of March 2026, that number is 67,665. A 78% increase from the bottom.

"AI Engineer" is the fastest-growing job title in the country. Postings up 143% year over year. AI's share of the tech job market went from 10% to 50% between 2023 and 2025. These roles carry a 56% wage premium. A year ago that premium was 25%.

5.6 million new business applications were filed in 2025. Double the pre-pandemic average. Five consecutive years at record levels.

More businesses means more work. Much of it does not have a name yet. The pie is expanding faster than it has in two decades.

The jobs exist. The system that connects you to them does not.

I know this because I hear it from candidates every day. Applications go into a void. 60% of listings on major platforms are reposts. AI tools that promise to help you apply are making it worse, because ATS systems now reject the automated applications those tools generate. One candidate told me he got callbacks on 2 out of 20 keyword matches using a handwritten resume, and complete silence with a 15 out of 15 AI-optimized version.

Every company in hiring builds for the employer. Nobody builds for the person looking for work.

When I started Flows, I could have built for companies. Hiring is where the money is, and every company in this space figured that out before I did. That's why every tool in hiring today is built for the employer.

The hard part, the part nobody is doing, is taking the tools that have only ever belonged to hiring managers and putting them in the hands of the people actually looking for work. Companies have AI that screens thousands of resumes in seconds. They have software that surfaces the best candidates and tracks every interaction. Why does the person on the other side of the table get a Workday form and silence?

I want to democratize that power. The same intelligence that companies use to filter people out should help people get in. A candidate with the right tools can find a role in days instead of months. I'm building toward that world.

The mission is to get a billion people hired.

The infrastructure that connects people to work will change significantly. It will create millions of new jobs that don't have names yet. The way people find those jobs will be different from how they do it now.

The "spray your resume across a hundred listings and hope one of them sees you" model is dead. Job boards won't replace it. Recruiters won't replace it. Nothing has replaced it yet, which is why people haven't noticed.

What replaces it is agents. Software that knows your work and your reality. Software that sits on your side of the table.

The search for the next chapter of someone's career should take days, not months. The signal between candidate and company should be honest, not optimized for keywords. The waiting should end.

Every technological revolution in the last two centuries has generated new work even as it eliminated old jobs. AI will likely follow this pattern and create more jobs than it destroys. But that depends on whether we build systems to match people with that work. Without them, billions will be unable to access opportunities because they're trapped behind broken application processes and lost resumes.