Francesco Ciccolella
Entrepreneurship in the Age of AI
Manish Kothari, MBA ’90, and professor Stefan Hepp, MBA ’90, discuss what it takes to build with technology changing the landscape.
- By
- October 07, 2026
- Entrepreneurship
Stefan Hepp, MBA ’90, adjunct assistant professor of entrepreneurship, and Manish Kothari, MBA ’90, met more than 35 years ago when they both lived at the university’s International House and attended Booth. Since then, Hepp has worked in institutional investing in Europe. He founded an asset-management company, led the firm for many years, and published a guide to private capital investing. Meanwhile, in Silicon Valley, Kothari has built companies—he’s taken one public and grown another, an app, to more than 100 million users—and backed more than 50 startups.
The two friends bring different vantage points to the same world. Here, they talk about the entrepreneurial journey and where the real opportunities for startups are now that AI is rewriting the rules.
Stefan Hepp: When you came to California a few years after Booth to work on your first startup with your brother, you slept on his couch. After more than a decade, you took that company public, and it ultimately got acquired by another public company.
Nowadays, you’re involved in a lot of startups: investing in them, talking to founders, coaching them on occasion. A generation ago, becoming an entrepreneur meant sacrifice; today it’s an alternative career path. What does that shift do to the mindset of younger founders, and how has the startup environment changed over all those years?
Manish Kothari: This doesn’t pertain to all of them, but for a lot of entrepreneurs, the viewpoint has shifted. There seems to me a little more obsession with valuations and exits than there used to be two or three decades ago. The ethos is a bit more about how much money you’re raising. Organizations like Y Combinator and others do a lot of good—they provide capital and a great network—but sometimes you’re left thinking, “For a seed-stage company valued at $20 million or $25 million, is that number really justified?”
For some AI startups, it probably is, because their product is deeply technical or requires a high spend on GPUs. But whether that amount of spending translates into business value and profitability remains to be seen for a large number of AI startups. Entrepreneurs have to balance growth with profligate spending.
We’ll find out in the next few years what all this money is doing for the AI space.
I always tell entrepreneurs: Focus on building a business. You don’t start with thinking about an exit or IPO. You start with, “I’m excited about this. I believe I have some edge. And I have the drive to continue with this, even when things go badly”—and they will at some point. Ideally, you need a growing market, conviction, a good team, and some luck.
“You don’t start with thinking about an exit or IPO. You start with, ‘I’m excited about this.’”
— Manish Kothari
Hepp: I’ve been thinking and writing a bit on the topic of AI financing, and as you know, I wrote a book about private capital investing. To my mind, the AI disruption story has two elements.
One element is the fear that it may upend traditional software-as-a-service or other software companies, which, from investors’ point of view, means that whatever the prospect of the new theme is, their money is stuck in the previous paradigm. They’re left wondering, “How much am I at risk of having backed the wrong horse?”
Kothari: Yes, the SaaSpocalypse, as it’s famously been called. There is some validity to that concern. If you look at a traditional B2B SaaS business such as Salesforce, its fundamental model for the past 20 years has been pricing on a per-seat basis, and that has worked well for them.
Because of AI, companies may no longer have as many users—seats to fill—as they once did, and so the pricing may shift from per seat to per AI token, which may negatively affect revenue and profit. But those companies still have distribution, which doesn’t go away. Salesforce and others have long-standing relationships with Fortune 500 companies and other large institutions, and that will be hard for startups to displace. It’s not that it cannot be done, but there are some real fundamental strengths to that relationship that we shouldn’t underestimate.
Additionally, these large firms have the proprietary data, complex workflows, and sophisticated compliance, regulatory, and security needs that make some of these SaaS companies well suited to take advantage of opportunities in AI.
Hepp: But they probably won’t do it all in-house. They’ll acquire expertise. This represents opportunity fields for startups. Meanwhile, frontier labs like OpenAI, Google, and Anthropic are moving up the stack in a quest to generate revenue. They want to be the input providers that can be used by companies needing a fenced wall around their data systems to develop further.
How does that affect the startup situation? What are some interesting prospects there?
Kothari: The frontier labs are competing fiercely with one another and with open-source models. They’re trying to grow their business by going into vertical spaces that they otherwise would not have—law, finance, education—and they’re making inroads.
Newer startups have a conundrum as to what spaces they can play in. They don’t have meaningful proprietary data, sophisticated compliance/security, or superior distribution. Nevertheless, I see a couple of areas of opportunity for them.
One place they can look at is what I would call old-world businesses: logistics, manufacturing, agriculture. The Midwest may have an advantage here in that it has such a diverse and diversified economy that there might be startup ideas where you can go to a logistics firm, a warehousing firm, or a manufacturing firm and come up with an AI solution and make it work for that particular segment.
Effectively, what I’m saying is, startups can behave as forward-deployed engineers, looking at areas where AI can really transform businesses. A forward-deployed engineer attempts to understand what a complex business might need in the future that they don’t know they need now.
I would broadly classify this avenue as doing something interesting with AI, married with a boring business. And I say “boring” not because it’s actually so—the business is usually thriving and profitable—but because you need to transform it into an AI-enabled one. And, by the way, that is not just a services play. Each manufacturing firm will have its own workflows, and the more complex and the more specific those workflows are, the more ideas there are for startups. Granted, these are not huge segments. But for that reason, OpenAI and Google aren’t going to dive into those areas for some time. They just don’t have the bandwidth to do it all.
The second place newer startups can look is in combining hardware and AI. Again, it’s not something that OpenAI or Google will do easily. That gives you a little bit of a moat. It’s an area I’m personally interested in because it hearkens back to my first startup, when my cofounders and I had an idea for something that was a new market category and became a case of classic disruption.
It’s about moving AI to something that is actually on a production line, on a conveyor belt, and that’s something even startup entrepreneurs can do. An example in this category would be a company that uses sensors with tiny, perhaps open-source models to do specific things, such as detecting the presence of certain gases on a line in real time.
“If you look at the world now, everything is much more about data than it was three decades ago. The data-driven Booth approach has really stood the test of time.”
— Manish Kothari
Hepp: Do these trends change the profile of the types of founders you’re seeing? My observation is that they tend to be young—almost right out of school—without a lot of specific industry or consulting experience, and they tend to be focused on programming, coding, and technology. But what you described points more toward needing a certain understanding of a sector, of a vertical.
Kothari: There’s a bit of a barbell effect. Broadly, B2C startups tend to draw younger people. They understand that space better, especially how to interact with consumers. If it’s B2B, it helps to have someone who has some experience with that industry and its workflows and nuances.
But more and more, and this is what’s exciting to me, it can be the two: the folks at the cutting edge—because they’re doing things that are so different, and it’s moving so quickly that only they can keep pace—alongside those who have more experience in a specific area. That combination can be pretty productive. We’re going to see more teams with that mix in successful companies.
Hepp: This is such a quickly evolving area.
Kothari: Even being here in Silicon Valley, it’s hard to keep up with everything that’s going on. I can’t predict what’s going to happen with AI, but I can only say that to my mind, making connections, showing empathy, and building understanding across groups will be important.
Hepp: I’m thinking back 35 years ago, when I had a decision to make: Do I want to go to America to do an MBA? This may sound naive nowadays, but I had this notion of what it would be like studying in the US, and it was fulfilled at Booth.
At the time, I-House was for graduate students, and not just in business but in law, medicine, divinity, philosophy. And of course there were people from all over the world. It was one of those fundamental things at Chicago that really enhanced my experience. And part of that was meeting you, so thank you for your friendship all these years.
Kothari: Likewise for me. What struck me when I first met you is that not only did you excel at finance and economics, but you were curious about things beyond our studies.
If you look at the world now, everything is much more about data than it was three decades ago. The data-driven Booth approach has really stood the test of time.
But what I have come to greatly appreciate, too, about my time at the school is the diversity of thought. You could have different points of view, even rival thinkers—someone who believed in efficient markets and someone who believed in the behavioral side of things—both in the same department, researching, writing, arguing, debating. This is so relevant right now. These components, outside of business and economics and finance and statistics, are fundamental to building something successful.