Below is my Friday speech, inspired by this bet, and following this live podcast recording.
I want to tell you about a bet I have with Brian Brackeen.
I’ve known for Brian for more than 10 years, back when he founded a facial recognition startup in Miami. He’s built in machine learning longer still. He and his wife and partner, Candice, cofounded Ohio-based VC firm Lightship Capital, investments around AI. He organizes Black Tech Week. He’s got opinions.
Last year we debated whether AI would imminently replace software developers. Brian said yes—developer jobs would fall. I disagreed. We bet a cheesesteak, or more properly having to wait in line for one, deadline May 2026.
It’s May 2026. This morning we’re going to learn the answer together, and what it means either way.
But the reason we’re all rightly interested in AI’s effects is the uncertainty. Let‘s first focus on what we do know about our economy, and what all of us here focus on.
Why exactly do we focus on entrepreneurship? Entrepreneurship, the creation of new firms, matters for three reasons.
First: it drags invention to market. Discovery sits in labs until someone builds a company around it.
Second: it manufactures jobs. Net new job creation flows from young firms—not small or big, but NEW.
Third: it injects dynamism. Competition forces improvement and shifts resources from dying industries to emerging ones.
New business. Not small. Not big. NEW.
In my book, the wealth creation that entrepreneur advocates often cite is not really the priority. It’s an outcome. We must challenge our entrepreneurs: If you solve problems, create jobs and spark dynamism for the rest of us, we should applaud wealth creation. Otherwise, we owe our entrepreneurs nothing, but their next swing.
And Americans seem to overwhelmingly approve of this deal. Polling shows entrepreneur is a more widely respected title than CEO, police officer or priest.
Entrepreneurship is the American church.
So why would anyone be against this? Who are entrepreneurship’s enemies?
Enemy Number One: Incumbents.
Every successful company hardens into an incumbent. Every incumbent learns it’s cheaper to rig the game than win it. So they weaponize regulation. Licensing requirements. Compliance regimes that strangle small competitors. Lawsuits designed to bleed, not win.
It’s not conspiracy. It’s incentives. Mature companies optimize for stability. New companies ARE instability.
Enemy Number Two: Economic developers who fence sit.
Here’s a hot take for this room: no conversation about entrepreneur support should question where funding comes from. The real debate of our era is how do we communicate that entrepreneurship and access to it is the economic strategy we need.
But economic development is a conservative trade trained for linear outcomes. Building housing and energy infrastructure requires a clear inputs to get predictable outcomes. We are less able to predict entrepreneurial outcomes. So what we require is many high quality bets from many different perspectives.
That’s what each of us work toward. But we know what feels like a sensible, conservative strategy is to
Too many still shovel tax breaks to the biggest company promising the most jobs. That’s not economic development. That’s wealth transfer.
If your job is sparking new economic activity, subsidize only companies that wouldn’t exist without you. NEW companies. Not relocated. Not expanded. NEW.
Every dollar to an incumbent starves the entrepreneur who might disrupt them.
Enemy Number Three: Ourselves.
Sometimes we sabotage entrepreneurship. We mistake activity for progress. We celebrate pitch competitions and ribbon cuttings but skip the question: did the trajectory shift? We romanticize “hustle” when the real barriers are structural—capital, networks, customers, talent.
We mistake entrepreneurship for personality instead of teaching it as skill.
And we calcify. We build ecosystems, then defend them like incumbents. We pick favorites. We recycle speakers. We stop interrogating ourselves.
Back to the Brian Brackeen bet.
AI might be entrepreneurship’s greatest enemy or greatest ally. I don’t know which.
The optimistic case: AI demolishes barriers. Solo founders build what used to require teams.
The pessimistic case: AI concentrates power with whoever controls models, data, infrastructure. Incumbents accelerate. The bar for “good enough” rises beyond what underfunded companies can clear. Both futures exist. We decide which wins.
If incumbents shape AI policy, we get rules that strangle competitors.
If economic developers pick AI winners, they’ll pick the safest—meaning the most established. If we treat AI like magic, we’ll get conned.
But if we stay sharp—if we ask “who benefits?” and “who gets shut out?”—we have a shot.
Tonight, you’ll witness the Cira Centre transform into a video game. Giant Tetris. Giant Pong. We could have used AI to render slick animation. We could have hired a production company.
Instead: LED lights, simple games, human creativity. Making something new with tools anyone can access.
But the bet that matters—who gets to build the future—we’re deciding right now.
The data tells us what happened. Adam Scavette will tell us what it means.
Adam’s a senior economist at the Federal Reserve Bank of Philadelphia. He studies entrepreneurship, regional economies, and how policy shapes who wins.