The Experimentation Machine - Jeffrey Bussgang - GEL 2025

The Rise of the 10x Founder: How AI Is Transforming Entrepreneurship

Jeff Bussgang:

Thank you all for being here today. It's wonderful to follow that amazing panel and see some familiar faces—two of the three panelists were my students at Harvard. I always love coming to Babson and seeing this cross-pollination between our institutions.

Every time I walk into a room like this, I carry a powerful belief: someone sitting here right now will build a billion-dollar company within the next decade. It could be one of the panelists, it could be you, but I genuinely believe extraordinary outcomes are available to each of you, especially given the AI revolution we're experiencing.

Today, I want to share the essence of my new book, "The Experimentation Machine." It's not about building an AI company per se, but about becoming an AI-native founder—a "10x founder" who uses modern AI tools to be exponentially more effective through what Anna Sophia called the "experimentation mindset" and what Yanka beautifully described as "iteration, iteration, iteration."

Beyond Founding: The Value of Joining

My previous books covered different entrepreneurial paths. The first helps aspiring founders raise money from VCs, while the middle one focuses on startup joiners—employees number 2 through 200. There's great honor in joining early rather than founding. My first job after business school was as employee number 30 at a venture-backed startup during the early internet days. We eventually went public, and that journey gave me the skills to later become a founder.

Someone in this room might become that billion-dollar founder—or a billion-dollar joiner. Take Zack Kirkhorn, my former student from 2013. After HBS, he joined Tesla's finance department, eventually becoming CFO. By age 40, he was the CFO of a trillion-dollar market cap company. So you can either be a billion-dollar founder or a trillion-dollar joiner. That's what's in front of you.

The 10x Founder Phenomenon

Many of you in technology have heard of the mythical "10x developer," right?

[Turns to audience member] Gonzalo, what's a 10x developer?

[Nods to Gonzalo's response]

Exactly! Someone who's 10 times more productive than average, not just 20% or 50% better. In software development, modern cloud tools have made this possible, and these 10x developers are the people you want for your startups.

Well, there's a running joke in the venture capital world: AI won't replace founders anytime soon, but founders who use AI will rapidly replace those who don't. Your challenge is to become a 10x founder, a 10x joiner, a 10x analyst—leveraging AI tools to be exponentially more effective as a leader.

The Greatest Wave Yet

Connor made a great point about technology waves creating opportunity and disruption. I've had the privilege of being part of the Web 1.0 era as an entrepreneur and investor, followed by Web 2.0, cloud computing, crypto, and mobile. But I can tell you with absolute certainty: the AI era is vastly more impactful than all of these combined.

[Shows chart on screen]

This chart tells the story simply. The x-axis is cost, the y-axis is capabilities. Look at how AI models are simultaneously getting dramatically cheaper while rapidly improving in capabilities. This isn't just Moore's Law times ten—it's an entirely different trajectory. Remember this: the AI you use today is the worst AI you'll ever use.

And it's everywhere—delivered for free on 8 billion smartphones, 2 billion laptops and tablets, and 1.5 billion cars. The speed of adoption is unprecedented.

Look at the top of this chart—"Human PhD." The latest models have capabilities matching PhDs in specialized fields. Scientific discovery is being enabled because AI knows the laws of physics, chemistry, and biology. This is driving breakthroughs in personalized medicine, chemical compounds, and even AI itself, creating a recursive loop that's bringing us closer to superintelligence.

The value creation opportunity? PWC estimates $16 trillion, equivalent to China's entire GDP. That's how much value will emerge from productivity improvements and new capabilities.

The New Normal

Unlike previous technologies with slow adoption curves, AI is becoming ubiquitous through your existing devices. Just as you take water, electricity, and good Wi-Fi for granted—[looks at an administrator in the audience] Provost Armony, I hope the Babson Wi-Fi is good!—you're already taking AI for granted in your daily life.

Yenko made a great analogy about eating oysters: the first time seems crazy, but soon everyone's doing it. Remember your first "aha" moment with AI? Maybe it was voice interaction, ChatGPT planning your vacation, or even using Uber. These once-revolutionary capabilities quickly became commonplace.

The Transformation Is Already Here

The companies at the forefront are already "eating their dog food." Google recently announced that one in four lines of its code is written by AI. Three months ago, that was. Today, I estimate it's over 30%, and soon it will be more than half.

Amazon's Andy Jassy, who's visiting HBS next week, described their AI platform for managing product data as saving the equivalent of 4,500 developer years. And in parentheses, he noted, "Yes, that number is crazy but real."

The CEO of OpenTable and Booking.com told me that 75% of customer service calls at OpenTable are now handled by AI agents. These companies are seeing 20-40% revenue growth with flat headcount, banking these productivity improvements and achieving more with less.

The Experimentation Machine

Mark Zuckerberg recently gave a fascinating interview—not the Joe Rogan one, I don't recommend that—but on the Acquired podcast. When asked about Meta's product strategy, his answer was exactly what Yinka just said. Zuckerberg said: "We don't have a product strategy. Our strategy is to iterate. If we iterate and learn as quickly as possible, we'll learn faster than any other company."

That's why I named my book "The Experimentation Machine." Your startup is essentially that—a learning machine designed to iterate as quickly as possible. My thesis is that AI catalyzes this iteration process, making it exponentially more efficient.

The Agentic Era

Let's do a quick survey. How many of you use ChatGPT weekly?

[Almost everyone raises their hands]

How many use Perplexity?

[Several hands go up]

Getting more sophisticated now—how many use Cursor, Lovable, or V0 as development tools to leverage English as a programming language?

[Two hands go up]

Just two? That's about to change dramatically.

I tell my MBA students we're in an extraordinary moment. It used to be that starting a company required a technical co-founder—sometimes dismissively called a "coding monkey." My Harvard students would trek to MIT, saying, "I've got this big idea, I just need someone to code it."

But now, software development has entered the "agentic era," where agents do the work, make decisions, and take actions for you. Tools like Cursor, Lovable, V0, and Replit are making it possible for non-technical people to build applications.

Anyone in this room—yes, even if you're non-technical—can build an app. What you need isn't coding skills but the ability to communicate effectively with these agentic tools, applying human judgment, strategic thinking, and insight.

I'm telling everyone in this room: you should all be builders. Six months from now, I bet half or two-thirds of you will be using these software development tools to build apps. That's how rapidly adoption is happening.

The Expanding Opportunity

This agentic future dramatically expands opportunities for entrepreneurs. One of my Flybridge colleagues analyzed just payroll automation—a $260 billion market. But when she calculated how AI agents could automate tasks across 170 vertical industries, the opportunity ballooned to $4 trillion—16 times larger.

As a tech entrepreneur, you're no longer just targeting IT budgets—you're going after entire operational budgets. When Connor thinks about his legal AI startup for diligence documentation, he's not targeting law firms' IT spending. He's replacing $300,000-a-year associates and their $1,000-an-hour billings.

The New Startup Model

The traditional venture model I grew up with is changing rapidly. At Flybridge, we'd typically invest $2-4 million in a seed-stage company that would hire 10-15 employees and aim for $1 million in annual recurring revenue before Series A. Their first move was always building a development team—hiring engineers from MIT, Northeastern, or Olin.

[Looks around] Any Olin College folks here? No, because they're already starting companies!

Today's model is completely different: you're the builder, the founder, closest to customer insight. You can achieve those same milestones with just 2-3 people plus a host of AI agents. In fact, within a few years, organizations—including Babson, nonprofits, and companies—will employ more AI agents than humans.

As a leader in this AI-forward world, you'll need new skills: spinning up digital employees, managing them, orchestrating their work, authenticating them, evaluating them, and ensuring they interface properly with humans. The management science we've developed for centuries at institutions like Babson and Harvard now needs to evolve to manage these agents.

The Solo Billion-Dollar Founder

Sam Altman from OpenAI shared something fascinating at a JP Morgan conference last year. He and his founder friends have a running bet: when will we see the first billion-dollar solo entrepreneur? When I mentioned there might be a billion-dollar entrepreneur in this room—maybe that's you, the solo founder.

An Andreessen Horowitz partner recently said, "I've lost count of the number of startups that have gone from 0 to $10 million in revenue in the last 12 months thanks to AI." The old startup mantra was "triple, triple, double, double"—from $1M to $3M to $9M to $18M to $36M over 5-7 years. Today, AI is compressing that timeline dramatically.

Bringing It All Together

The essence of my message is this: leverage modern AI tools to become a 10x founder, but combine them with timeless principles of strategy, customer discovery, value creation, and building enduring, profitable companies. You need both elements—you can't just master the tools without understanding value creation, nor can you focus solely on business fundamentals without embracing AI.

Let me leave you with a quick example: Topline Pro, founded by two HBS dropouts (in the tradition of Gates and Zuckerberg). They built a platform for service professionals—landscapers, electricians, painters—that automatically creates websites and manages customer relationships. As each new AI model emerged—GPT-2, 3, 4, and now 4.5—their platform grew more powerful.

They're transforming how service professionals—who prefer hands-on work to administrative tasks—manage their businesses. And as AI capabilities become recursive, as I mentioned earlier, we're approaching a point where AI is so powerful it's researching and improving itself.

That's the opportunity in front of you. Who among you will seize it?

[Looks around the room with anticipation]

Thank you.